The Role of Extracellular ATP in Dictyostelium discoideum Growth and Development

Transcription

1 The Role of Extracellular ATP in Dictyostelium discoideum Growth and Development Peter F. Woolman A Thesis Presented to School of Biological Sciences, University of East Anglia In Fulfilment of the Requirements of the University of East Anglia for the Degree of MSci By Research December 2013 This copy of thesis has been supplied on the condition that anyone who consults it is understood to recognise that its copyright rests with the author and that no quotation from the thesis, nor any information derived from it, may be published without the author s prior written consent. 1

2 Declaration I hereby declare that the work in this thesis is my own work and effort and that it has not been submitted anywhere for any award. Where other sources of information have been used, they have been acknowledged. Signature:. Date:. 2

3 Abstract It has long been known that the social amoeba Dictyostelium discoideum initiates and controls its developmental processes by the creation and transmission of the purinergic molecule; cyclic AMP. It is also known that it possesses P2X receptors which are localised on the inside of the cell. Despite the lack of cell surface ATP receptors, previous studies have indicated the existence of Dictyostelium ecto-atpases and calcium responses to the addition of exogenous ATP. In these experiments it was found, using the luciferase assay, that the cells condition their media with ATP during axenic growth and development. Treatment of cells with apyrase, a potato enzyme known to degrade ATP, did not affect the rate of growth, but led to a delay in the appearance of streams during development. This delay corresponded with the formation of clumps of cells, but no evidence could be found that the cells had increased expression of adhesion molecules. Apyrase also caused developing cells to form fewer fruiting bodies, despite the fact that there was no difference in the number of aggregates formed, implying that the apyrase was interfering in the transition between these stages. The apyrase fruiting bodies also contained fewer viable cells, though whether it was fewer cells in general or less healthy ones could not be confirmed. The malachite green assay was used to measure the rate of activity of the cell surface ecto-atpases and the results suggested that the common ecto-atpase inhibitor suramin would affect them. Cells allowed to develop in the presence of suramin showed no abnormal timing in the stages of development or in number of fruiting bodies, but did contain significantly more viable cells. Altogether these results are consistent with the hypothesis that ATP, ATP receptors and ATPases play a role in the regulation and co-ordination of development in Dictyostelium discoideum, despite the fact that it possesses no known cell surface P2X receptors. 3

4 4 Table Of Contents Abstract... 3 Table Of Contents... 4 List Of Figures... 8 List Of Abbreviations Acknowledgements Chapter 1 - Introduction Purinergic Signalling Purinergic Signalling - History Purinergic Signalling Mechanisms Of ATP Export From Cells Purinergic Signalling Mechanisms Of Extracellular ATP Detection Purinergic Signalling ATP Breakdown Dictyostelium discoideum Dictyostelium discoideum - Background Dictyostelium discoideum Development There Are Potentially Further Purinergic Signalling Systems Regulating Development In Dictyostelium discoideum Chapter 2 - Materials And Methods Cell Culture Use Of The Haemocytomer Measuring Growth Curves Development Buffer Development In Submerged Culture Treatment With Apyrase Treatment With Heat-Inactivated Apyrase Development On Agar Determining Fruiting Body Size Measuring Extracellular ATP Measuring Cell Adhesion Aggregation Quotient Measuring Adhesion Measuring ATPase Activity Phosphate Liberation Buffer MLG Reagent... 27

5 Molybdate Solution Measuring Phosphate Liberation Data Analysis Chapter 3 Investigating The Presence Of Extracellular ATP And The Effect Of Its Removal Upon Growth And Development In Dictyostelium discoideum Aim Introduction Dictyostelium discoideum Undergoing Exponential Growth Causes Luciferase To Produce More Luminescence Than Untreated Buffer The Addition Of Exogenous Apyrase Has No Effect On Exponential Growth Of Dictyostelium discoideum Dictyostelium discoideum Secretes ATP During Development Submerged Culture Submerged Culture - Characterising The Control Submerged Culture The Addition Of Exogenous Apyrase To Dictyostelium Cells Causes A Developmental Delay Submerged Culture The Addition Of Exogenous Apyrase Impacts The Development Of Dictyostelium Cells That Are Already Undergoing Development Submerged Culture Heat Inactivated Apyrase Is Incapable Of Causing The Apyrase Induced Delay Submerged Culture The Addition of Extracellular Adenosine Deaminase To Dictyostelium Does Not Counteract the Apyrase Induced Delay In Stream Formation Submerged Culture The Effect Of Exogenous Apyrase On Dictyostelium Development Is Not Limited To The AX4 Strain Cell Adhesion Cell Adhesion Characterising The Control Cell Adhesion Addition Of Exogenous Apyrase To Dictyostelium Does Not Alter the Aggregation Quotient Development On Agar Development On Agar The Addition Of Exogenous Apyrase Induces A Developmental Delay In Dictyostelium discoideum Developing On Agar Development On Agar The Addition of Exogenous Apyrase Decreases the Number of Fruiting Bodies formed by Dictyostelium discoideum Development On Agar The Addition Of Exogenous Apyrase Decreases the Number Of Viable Cells Within Fruiting Bodies Produced By Dictyostelium discoideum Development On Agar Investigating Whether There Is A Link Between Number Of Fruiting Bodies And Number Of cells Within The Fruiting Body Heads Conclusions... 53

6 Chapter 4 Investigating Ecto-ATPases And Their Role In Regulating Development In Dictyostelium discoideum Aim Introduction Phosphate Liberation Phosphate Liberation Measuring Change In Absorbance In Control Conditions Phosphate Liberation Measuring Change Of Absorbance In The Presence of Suramin ATPase Activity Measuring Change Of Absorbance In The Absence Of Mg 2+ Ions The Effect Of Suramin On Dictyostelium Development In Submerged Culture The Effect Of Suramin On Dictyostelium Development On Agar Conclusions Chapter 5 - Bioinformatics Aim Introduction Overview Of The Process Used Conserved Sequences Protein BLAST Search Transmembrane Alignment Identifying Domains Results Results - DDB_G Results Other genes Conclusions Chapter 6 - Discussion An Overview Of The Findings Of This Project Dictyostelium discoideum Responds To Increased Degradation Of Extracellular ATP Addition Of Apyrase To Dictyostelium Causes A Delay In The Initiation Of Streaming And The Formation Of Clumps Alterations In The Rate Of Breakdown Of Extracellular ATP Causes A Change In The Number Of Viable Cells Making Up Dictyostelium discoideum Fruiting Bodies The Rate Of ATP Degradation May Affect The Number Of Cells Making Up A Dictyostelium discoideum Fruiting Body The Rate Of ATP Degradation May Affect The Viability Of Cells Making Up A Dictyostelium discoideum Fruiting Body Addition Of Apyrase To Dictyostelium discoideum Causes A Decrease In The Number Of Fruiting Bodies Formed

7 6.6 What Further Work Could Be Done To Answer The Questions Raised By This Project? References Appendix - Dictyostelium Genes Identified By The BLAST Analysis That Had Been Determined To Contain Two Transmembrane Domains And The Correct Topological Alignment

8 8 List Of Figures Figure 1 Use Of The Haemocytometer. a) The view of the counting grid of a haemocytometer loaded with Dictyostelium discoideum cells b) the same haemocytometer with the squares to be counted labelled Figure 2 Increased Luminescence From The Luciferase Reaction In Culture Medium Conditioned by Growing Dictyostelium. The luciferase-luciferin reaction results in significantly less luminescence in unconditioned FM minimal growth media than media conditioned with growing Dictyostelium (5x10 5 cells/ml; 17 hours at 125RPM). The error bars represent standard error of the mean of each data point, N=4, P< Figure 3 Growth Curves Of Dictyostelium. Growth over a period of 54 hours in FM minimal media under control conditions or in the presence of 4U/ml apyrase. The error bars represent standard error of the mean of each data point, N=4, P> Figure 4 - Detection Of ATP Upon Addition Of Dictyostelium To Development Buffer. Mean ATP detected by luciferase-luciferin reaction (luminescence, RLU) over the 120 minutes following seeding of Dictyostelium into development buffer at 5x10 5 cells/ml. Each data point is subtracted from the value obtained at the same time point from buffer that did not contain cells. The error bars represent standard error of the mean of each data point Figure 5 Development In Submerged Culture. Dictyostelium development in submerged culture at cell density of 1x10 6 cells/ml. Photos taken at 4, 7, and 24 hours after addition to the wells. Images are annotated with descriptions of the developmental process Figure 6 - Developmental Progress Of Dictyostelium After 14 Hours At 4 Different Cell Densities (cells/ml). Aggregates of cells are visible at densities of 1x10 6 cells/ml and 5x10 5 cells/ml Figure 7 Development Of Dictyostelium In Submerged Culture At Cell Density Of 5x10 5 cells/ml. Between hours cells begin to elongate and lose their amoeboid shape. At 18 hours the elongate cells begin to sort themselves into patterns and very early streams, only a couple of cells long, can be seen. By 20 hours the streaming is occurring, the cells are forming a vortex shape as they determine where the eventual aggregation centres will be. These aggregation centres are visible at 22 hours and the cells are moving inwards along the streams towards them. By 24 hours the streams have begun to break up, each segment will form their own separate aggregate Figure 8 Apyrase Causes A Delay In The Formation Of Streams. Development of AX4 cells at a density of 5x10 5 cells/ml in cultures treated with 4U/ml of apyrase or an equivalent volume of water over a period of 24 hours. Time since addition to the buffer is given in hours on the left of the image. Photos are of the most developmentally advanced cells found within the well. In control wells streams are first visible at 20 hours and begin to break up by 24 hours. The Apyrase-treated cells, however, take a further two hours before they start to form streams Figure 9 No Effect Of Apyrase On Complete Aggregate Formation. After 30 hours in development buffer, aggregates of control Dictyostelium cells at a density of 5x10 5 cells/ml do not have observable differences to those formed in the presence of 4U/ml apyrase Figure 10 Apyrase Causes The Formation Of Clumps. After treatment with 4U/ml apyrase, Dictyostelium at a density of 5x10 5 cells/ml form "Clumps." Images taken 1.5 hours after additon of cells to the buffer Figure 11 - Evidence For Increased Incidence Of Cells Failing To Commit To Aggregate Formation Upon Treatment With 4U/Ml Apyrase. Cells at 5x10 5 cells/ml. Images taken at 20hours after addition to the buffer Figure 12 Apyrase Induced Developmental Delays Caused No Significant Difference In The Eventual Number Of Aggregates Formed. Number of aggregates present in wells containing control

9 and apyrase (4U/ml) treated cells following completion of development at two different densities of Dictyostelium. The error bars represent standard error of the mean of each data point, N=7. Each data point is expressed as a percentage of the average amount of aggregates within each pair s control well Figure 13 Streaming Is Halted By Addition Of Apyrase. Apyrase (4U/ml) was added to Dictyostelium cells at a density of 5x10 5 cell/ml half an hour after the starting of stream formation. Time in hours since addition of cells to the buffer is recorded in the top left corner of each image Figure 14 No Observed Effect Of Heat-Inactivated Apyrase On Development. Cells after 24 hours at 5x10 5 cells/ml under control conditions or with apyrase (4U/ml) that was either enzymatically active or heat inactivated Figure 15 Adenosine Deaminase Has No Effect On The Timing Of Dictyostelium Stream Formation. Photo taken 13 hours after addition to development buffer. In the wells that do not contain apyrase the cells are forming streams. In the wells that do contain apyrase there are no streams present but cells are forming clumps. Cells at a density of 5x10 5 cells/ml under control conditions or with apyrase (4U/ml) and with or without adenosine deaminase (2U/ml) Figure 16 Additive Effect Of Apyrase And Adenosine Deaminase On Number Of Aggregates Formed During Development. Number of aggregates present in wells seeded with 5x10 5 cells/ml Dictyostelium under control conditions or treated with apyrase (4U/ml), adenosine deaminase (2U/ml) or both proteins after a period of 42 hours. The error bars represent standard error of the mean of each data point, N=3 with two wells in each experiment Figure 17 - Apyrase Causes A Delay In The Initiation Of Streaming Of AX2 Cells. Development of AX2 cells seeded at a density of 5x10 5 cells/ml in cultures treated with 4U/ml of apyrase or an equivalent volume of water over a period of 24 hours. Time since seeding is given in hours on the left of the image. Photos are of the most developmentally advanced cells found within the well Figure 18 Aggregation Quotient Changes Over Time Under Control Conditions. Dictyostelium cells at a density of 5x10 5 cells/ml adhere to each other when mixed in the buffer through chance collisions and the expression of adhesion molecules a) Mean Aggregation Quotient over a period of 4.5 hours. The error bars represent standard error of the mean of each data point, N=8. b) Representative figure of the change of aggregation quotient in one tube over a period of 4.5 hours Figure 19 Addition Of Apyrase Has No Effect Upon The Aggregation Quotient. The change in Aggregation Quotient over time of Dictyostelium at a density of 5x10 5 cells/ml is not altered when apyrase (4U/ml) is added. The Aggregation Quotient is, however, altered by addition of EDTA (5mM) which prevents cell-cell adhesion. Control N=15, apyrase N=9, EDTA N=6. The error bars represent standard error of the mean of each data point Figure 20 - EDTA Prevents Development. Dictyostelium seeded at 5x10 5 cells in submerged culture over a period of 20 hours either with or without exogenous EDTA (5mM). Control cells start to form streams at 15 hours and by 20 hours are beginning to become complete aggregates. EDTA treated cells fail to show any signs of development Figure 21 Rounded Appearance of EDTA Treated Cells. Within 1.5 hours of being added to the submerged culture, Dictyostelium seeded at 5x10 5 cells/ml and treated with EDTA (5mM) are noticeably rounder than ones in control conditions Figure 22 Apyrase Causes Delayed Development On Solid Agar. Development of AX4 cells seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar. The cells treated with 4U/ml water are still at the mound and slug stages of development after 22 hours. At this time the control cells are mostly at the fruiting body stage with some remaining slugs Figure 23 Apyrase Causes A Decrease In The Number Of Fruiting Bodies Formed. Mean number of fruiting bodies formed by Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient 9

10 agar with or without apyrase (4U/well). Each data point is expressed as a percentage of the relevant control. The error bar represent standard error of the mean the data, N=6, two wells/experiment Figure 24 - Evidence A For Smaller Number Of Viable Cells In Fruiting Bodies Developed During Apyrase Treatment. Number of colonies produced by 0.5ml of the mixture of a fruiting body head and 4ml of Klebsiella. Spores were produced from by Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar under control conditions or in the presence of 4U/well of apyrase. Each data point is expressed as a percentage of the relevant control. The error bars represent standard error of the mean the data, N=24 plates of control and 28 of apyrase Figure 25 - The Number Of Fruiting Bodies In A Well Containing 2.5x10 6 Cells Plotted Against The Mean Number Of Colonies Produced By Said Fruiting Bodies Figure 26 There Is Significant Change In The Absorption Of light From Samples Taken From Wells Containing Cells After Addition Of Exogenous ATP Between Time Zero And Either Three Minutes Or Six Minutes. Data at three minutes is significantly different in each of the samples from wells containing cells from the control wells at the same time point (p<0.05). There is no statistical difference at six minutes between the wells containing cells and the control that do not, but the wells containing cells still have significantly more absorbance (p<0.05) than the same well at time zero. Data points taken at 3 and 6 minutes. The error bars represent standard error of the mean of each data point, N= Figure 27 In The Presence Of Exogenous Suramin, Exogenous ATP Causes No Alteration In Absorbance At 600nm In The Six Minutes After Addition. Addition of ATP (0.5mM) in the presence of suramin (250µM) causes no significant difference in absorbance of light at a wavelength of 600nm. Data points taken at 3 and 6 minutes. The error bars represent standard error of the mean of each data point, N= Figure 28 Removal Of Mg 2+ Ions From The Reaction Buffer Decreases The Rate Of Ecto-ATPase Activity. Addition of ATP (0.5mM) with MgCl left out of the reaction buffer causes a significant difference in absorbance of light at a wavelength of 600nm between samples taken from the wells containing cells at time zero and the same wells at six minutes. No significant difference was observed in the absorbance of light between samples taken from wells at time zero and three minutes in any of the wells. Data points taken at 0,3 and 6 minutes. The error bars represent standard error of the mean of each data point, N= Figure 29 No Effect Of Suramin On Development. Development of AX4 cells at a density of 5x10 5 cells/ml in cultures treated with 250µM of suramin or an equivalent volume of water over a period of 24 hours. Cells start to form early streams at 5.5 hours and by 6.5 hours streaming is fully underway. By 24 hours aggregates have formed. Time is given in hours on the left of the image. Images are of the most developmentally advanced cells found within the well Figure 30 Treatment With Suramin Caused No Significant Difference In The Number Of Aggregates Formed By Dictyostelium. Number of aggregates present in wells containing control or suramin (250µM) treated Dictyostelium seeded at 5X10 5 cells/ml following completion of development. The error bar represents standard error of the mean of the data, N=3, 2 wells/experiment Figure 31 Suramin Has No Effect On The Timing Of Dictyostelium Discoideum Development On Agar. Development of AX4 cells seeded at a density of 1x10 4 cells/cm2 on solid non nutrient agar treated with 250µM of suramin or an equivalent volume of molecular-grade water at 19 and 43 hours after addition to the plate. At 19 hours cells in all the wells had formed mounds. By 43 hours all the plates contained fruiting bodies Figure 32 - Suramin Has No Effect On The Number Of Fruiting Bodies Formed. Mean number of fruiting bodies formed by Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient 10

11 agar with or without suramin (250µM). Each data point is expressed as a percentage of the number of fruiting bodies in the paired control well. The error bar represents standard error of the mean the data set, N=7, two wells/experiment Figure 33 - Evidence For Increased Number Of Viable Cells In Fruiting Bodies Developed In The Presence Of Suramin. Number of colonies produced by 0.5ml of the mixture of a fruiting body head and 4ml of Klebsiella when grown upon an agar plates. Spores produced from Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar with or without suramin (250µM). Each data point is expressed as a percentage of the number of colonies produced by the relevant control. N=29 plates of control and 21 of suramin. The error bar represents standard error of the mean the data set. 67 Figure 34 - Sections Of The Clustalw Output (Larkin et al., 2007) With The Regions That Were Highly Conserved Marked And Numbered Figure 35 TMHMM Output For Four Proteins (Analysis, 2007). a) Protein gi_ has four transmembrane domains, which is too many so it is discarded. b) Protein gi_ has no transmembrane domains. It cannot be anchored in the membrane and so it is discarded. c) Protein gi_ contains two transmembrane regions, but the segment between them is located on the inside of the cell membrane. Even if this protein has ATPase activity, it would be acting on nucleotides on the inside of the cell and so is irrelevant to this investigation. It is discarded. d) Protein gi_ has two transmembrane domains separated by an external region. This is carried on for further investigation Figure 36 - TMHMM Output For Gene DDB_G (Analysis, 2007)

12 List Of Abbreviations This list provides clarification to some of the abbreviations that are commonly used throughout this thesis ADP Adenosine DiPhosphate AMP Adenosine MonoPhosphate ATP Adenosine TriPhosphate cada Gene that expresses csb protein csb Contact Site B adhesion molecule camp Cyclic Adenosine MonoPhosphate EDTA - Ethylenediaminetetraacetic Acid EGF Epidermal Growth Factor EGTA - Ethylene Glycol Tetraacetic acid E-NTPDase - Ecto-Nucleoside TriPhosphate Diphosphohydrolase HL5 A media for axenic Dictyostelium discoideum growth composed of yeast FM Minimal A media for axenic Dictyostelium discoideum growth containing no yeast MLG Malachite Green P1 A class of Adenosine detecting Purinergic Receptor P2X A class of Purinergic ATP Receptor P2Y A class of Purinergic Receptor to ATP and ADP PDE Phosphodiesterase PSF Prestarvation Factor UTP - Uridine TriPhosphate 12

13 Acknowledgements First and foremost thanks must go to Dr Samuel Fountain, who gave me the opportunity to work in his laboratory, carry out this project and always had helpful recommendations. I also thank the other members of the Fountain Lab: Dr Venketesh Sivaramakrishnan, who introduced me to the tenets of Dictyostelium culture and development, and Hinnah Campwala, who kept me company in the lab and provided feedback on my work at weekly lab meetings. I am also deeply indebted to my secondary supervisor, Professor Tracey Chapman, for helping me obtain this project and for providing advice and guidance whenever I lost track of the bigger picture. I would like to thank my parents who raised me, encouraged me to pursue a career in research and provided financial support during this project. Thanks also go to the students of the Pamela Salter Room, who guided me in the ways of Post-Graduate life, and Florence Morgan-Richards, who provided essential emotional support. Finally I thank my Grandfather, Dr John Peter Whitehouse, who regaled me with tales of scientific curiosities in my youth and who inspired me to become a scientist. 13

14 Chapter 1 - Introduction 1.1 Purinergic Signalling Purinergic Signalling - History Traditionally neurotransmission has been thought of as being divided into the two categories cholinergic and adrenergic, with nerves releasing either noradrenaline or acetylcholine. It was not until the 1970s that the case was put forward for the existence of a third category, though, in retrospect hints as to its existence were seen as early as the late nineteenth century. This third category was tenatively called purinergic because, at the time, it was believed to use only a single purine nucleotide as the signal (Burnstock, 1972). Since these early investigations purinergic ATP signalling has been found in almost all nerves in both the central and peripheral nervous systems of mammals, while purines and pyramidines carry signals in virtually all human non-neuronal tissues (Burnstock and Verkhratsky, 2009). The ubiquitous role of ATP as the primary energy source of cells has given rise to the belief that it was selected for this task very early on in evolution and that, as a result, even the most primitive early cells would have had a high internal ATP concentration. When these cells died their degradation would have released this ATP, causing ATP gradients in the primordial seas. It is believed that this is what initially led to the development of the purinergic signalling systems. Cells evolved mechanisms to detect this gradient in order to avoid dangerous places where cell death was common (Burnstock and Verkhratsky, 2009). Over time ATP receptors have grown more elaborate with more specialised purposes. Its early appearance in evolution explain why ATP seems to be the main exception to the rule that extracellular signalling molecules are functionally and anatomically segregated. ATP is not only found in specific tissues doing specific tasks. Today it can be found in a wide variety of roles and tissues (Burnstock and Verkhratsky, 2009) including, but not limited to; acting as a neurotransmitter in smooth muscle (Mutafova-Yambolieva, 2012), modulating Ca 2+ signalling in too many different contexts to list here (Mammano, 2012), modulating the immune system in ways which are exploited by cancers (Pellegatti et al., 2008) and even maintaining homeostasis in bones (Ham and Evans, 2012). ATP also posseses a wide variety of release pathways. The widespread nature of ATP signalling implies that it appeared very early in evolution (Burnstock and Verkhratsky, 2009) and it is the wide variety of its functions which makes ATP-regulated purinergic signalling such an important field of study. Interestingly, while the ATP-sensitive P2X receptors are found in a wide range of organisms, from humans to amoeba, they are also absent in many organisms such as higher plants, fruit-flies and yeasts (Sivaramakrishnan and Fountain, 2012). The phylogeny of other ATP-sensing receptors, such as the P2Y receptor is less clear with reports concerning their expression in plants and simple organisms, including Dictyostelium, lacking Purinergic Signalling Mechanisms Of ATP Export From Cells There are several ways in which extracellular ATP can be released from the cell, the first of which has been alluded to already. If a cell can be induced to burst, such as in the event of osmotic cytolysis, it will cause the release of all its contents: including the 3-5mM ATP found within the cytoplasm under 14

15 normal, healthy circumstances (Beigi et al., 1999). This is a non-specific mechanism of ATP release. Extracellular ATP is released along with the rest of the cytoplasmic contents and, of course, it results in the death of the cell from which the ATP originates. This would rarely be a deliberate action by the cell, but it would provide an unintended warning signal to other cells in the vicinity. This is probably the initial mechanism from which the entire purinergic network has evolved (Burnstock and Verkhratsky, 2009). A second way in which ATP can be released from the cell is as part of a secretory vesicle. Proteins synthesised within the cell, but which need to be localised to the outside of the cell membrane, or even into the extracellular space, are packaged into membrane-enclosed containers called vesicles. These are then transported to and fuse with the cell s membrane, releasing the vesicle s contents on the outside of the cell (Sherwood, 2012). It is through this method that more traditional signalling molecules, such as noradrenaline in nerves, are released from the cell. It is now known that ATP can also be packaged inside these vesicles for release (Lohman et al., 2012). ATP Binding Cassette, or ABC, transporters are a type of membrane protein which move molecules across the plasma membrane which might otherwise not be able to do so due either to their polarity or diffusion gradients. The energy for this process is gained by the removal of a phosphate from ATP and thus each of these proteins possess two conserved ATP binding domains (Borst and Elferink, 2002). There is, however, a growing school of thought that suggests some of these transporters not only use ATP as an energy source for transport, but actually transport it as well. Such an idea is still controversial. As of the time of writing papers are still regularly being published with conflicting evidence on the subject. Some scientists even suggest that the ABC transporters do not release ATP themselves, but regulate an unknown ATP release mechanism (Lohman et al., 2012). The existence of ABC transporter-mediated ATP release is not confirmed. The conflicting evidence, however, means that it must be considered as a potential source of extracellular ATP in studies where it can be found to be present without a known release pathway. Traditionally ATP has been considered as being created inside cells. In mammals the ATP synthase protein has traditionally been believed to exist only within the mitochondria. In another example of advances in the field of purinergics overturning established knowledge of ATP, however, there is evidence to suggest that, in some forms of human tumour cells at least, ATP synthase homologous molecules exist on the outside of the cell membrane (Das et al., 1994). This suggests there are some cells capable of generating extracellular ATP from ADP directly in the extracellular environment without needing to export it Purinergic Signalling Mechanisms Of Extracellular ATP Detection There are nineteen types of purinergic receptors which are sorted into three classes: P1 receptors which respond to adenosine and the two forms of nucleotide triggered P2 receptors; G proteincoupled P2Y and ionotropic P2X (Mathieu, 2012). The P2X receptors are then further divided into P2X1-P2X7 subfamilies (Surprenant and North, 2009), while the P2Y receptors have eight subfamilies. Unlike the neat 1-7 nomenculature of the P2X receptors, the P2Y families consist of P2Y1, P2Y2, P2Y4, P2Y6, P2Y11, P2Y12, P2Y13, and P2Y14. The numeric gaps represent proteins that, while homologous to P2Y receptors, have not been shown to have any response to nucleotides. It was also only after the classes had been subdivided in this way that it was discovered that some P2Y receptors respond to pyrimidines instead of purines (Abbracchio et al., 2006). 15

16 P2X receptors are gated channels, which respond to ATP with a rapid change in permeability to allow Na +, K + or Ca 2+ cations to pass through. P2X7 is unique amongst them, in being the only one whose pore is capable of acting non-selectively (Ralevic and Burnstock, 1998). A P2X subunit consists of two transmembrane domains, named TM1 and TM2 (Jiang et al., 2012). The bulk of the protein consists of an extracellular hydrophilic loop of amino acids, which separates them. It is important to note that a subunit alone is incapable of forming an ion channel (Ralevic and Burnstock, 1998). One P2X receptor is believed to be composed of three subunits with each pore being made up of three separate TM2 domains surrounded by a ring of three TM1 domains (Jiang et al., 2012). Unlike the P2Y receptors (except P2Y11), P2X receptors contain introns in their coding sequence (Abbracchio et al., 2006). Sequence analysis of P2X receptors from various species reveals that the regions that are highly conserved, and which are essential for the function of the protein in mammals, are found mostly in the transmembrane domains (Surprenant and North, 2009). The vast bulk of the ecto-domain, which had previously been considered to be the crucial part, is barely conserved at all. There are four amino acids in the area immediately following the transmembrane domain: Tyr 55, Gln 56, Lys 69 and Lys 71. These are the only amino acids known to be vital for receptor activation. It is now believed that only the Lys residues are essential for ATP binding (Surprenant and North, 2009). Crystalography of the Zebra Fish P2X receptor suggests that, in this specific receptor at least, the binding site resembles an open jaw. When ATP binds to this, the jaw tightens and this conformational change in the protein shape opens up the pore (Jiang et al., 2012). P2Y receptors are rhodopsin like g-coupled proteins consisting of seven transmembrane domains (von Kügelgen and Wetter, 2000). These respond as much as ten times slower than P2X receptors to extracellular ATP (Ralevic and Burnstock, 1998). Each subfamily of P2Y receptor can couple to a wide range of different g-proteins in order to give a wide variety of different effects. It has been suggested that this variety of different g-proteins can itself give rise to specificity to which agonist the receptor binds. For instance P2Y11 receptors exposed to ATP will increase camp and Ca 2+ levels. When P2Y11 receptors are exposed to UTP, however, only Ca 2+ will be released, while the camp permeable receptors remain unactivated (Abbracchio et al., 2006) Purinergic Signalling ATP Breakdown Another important component of purinergic signalling systems are the ATPases which hydrolyse ATP. These are essential in that, not only do they terminate the signal activating the P2X receptors, but they also generate ADP, which activates the P2Y receptors. Furthermore, many ATPases are also capable of breaking down this ADP further into AMP (Knowles, 2011). In mammals CD73 proteins then convert this AMP into adenosine, which activates P1 receptors (Colgan et al., 2013). In this way one molecule of ATP can briefly activate many receptors as it is broken down (Knowles, 2011). The class of ATPases responsible for ATP hydrolysis in purinergic signalling is the ecto-atpase. This consists of membrane-bound proteins, which are insensitive to the main inhibitors of most other classes (Chadwick and Frischauf, 1997) and which require Ca 2+ or Mg 2+ ions to function (Kukulski et al., 2005). It was initially thought that only one ecto-atpase existed, named CD39, with a variety of post-translational modifications. As more were discovered, however, the existence of the entire ENTPDase (ecto-nucleoside triphosphate diphosphohydrolase) family was acknowledged (Robson et al., 2006). 16

17 All of the eight known ENTPDase proteins, except 5 and 6, are located on membranes with 1,2,3 and 8 being on the cell s external membrane. They are connected to the membrane by two transmembrane helices with the catayltic region on the outside of the cell in order to act upon the extracellular ATP. These transmembrane domains, and indeed their presence within the cell membrane, are critical for the catalytic ability, at least in the CD39 protein, as they hold the active site in specific shapes. It is believed that the variety of active site shapes caused by transmembrane anchoring is responsible for the variety of substrate specificities present in the ENTPDase family (Grinthal and Guidotti, 2006) Dictyostelium discoideum Dictyostelium discoideum - Background Dictyostelium discoidium is a model organism that is used in a wide variety of fields from studies of phagocytosis, to cell signalling, cell motility (Montagnes et al., 2012), cell adhesion and even evolution (Abedin and King, 2010). In 1998 an initiative to sequence the Dictyostelium discoideum genome was launched; the first attempt to sequence the full genome of a free living protozoan. The completed genome was published in Nature in 2005 (Eichinger et al., 2005), thus further increasing its popularity and usefulness as a model organism. Dictyostelium is of particular use in this study because it has been known since 2007 to contain P2X receptors (Fountain et al., 2007) and was thus confirmed to use ATP as a signalling molecule. Later research indicates that the known receptors are all positioned to detect ATP on the inside of the cell membrane and thus this signalling is all intracellular (Ludlow et al., 2009). Ludlow et al (2008), however, demonstrated that extracellular ATP could activate a hitherto unknown calcium influx pathway, suggesting that there is expression of cell surface ATP receptors in Dictyostelium. Cellular slime moulds are used in developmental studies due to the fact that they are amongst the simplest eukaryotic organisms capable of acting multicellularly (Bracco et al., 2000). Although this multicellularity is known to have evolved independently of the same phenotype in animals and plants, many of the same gene families were vital for both transitions (MacWilliams et al., 2006). Furthermore, while this multicellularity is achieved not through growth and differentiation of cells, but through aggregation (Dormann et al., 2000), it retains many of the same developmental pathways of the more difficult to study metozoans from which they diverged millenia ago (Coates and Harwood, 2001). During the early stages of their life cycle slime moulds, such as Dictyostelium discoidium, typically exist as solitary amoeba, reproducing via binary fission and feeding on bacteria found within their environment by phagocytosis (Bracco et al., 2000). Axenic strains of many have been created in the laboratory to ease the undertaking of genetic studies (Franke and Kessin, 1977). Upon starvation many amoeba in an area come together to form an aggregate. The size of this aggregate can vary from 100 to cells (Bracco et al., 2000). This aggregate differentiates into stalk and spore cells. The stalk cells form a hard structure that hold aloft the spore cells, which have formed a sorus, thus allowing them to disperse over a large area. The spore cells are picked up by the wind and transferred to a new location where there is likely to be more nutrients available than their previous resource depleted area. Here they can continue their unicellular life cycle, while the stalk cells remain at the old location and die (Kuzdzal-Fick et al., 2010). The altruistic nature of the life cycle of these cells seems counterintuitive from an evolutionary perspective. Dead cells do not get to reproduce. This is, in fact, a classic example of kin-selection 17

18 theory which states that cells (or higher organisms) will sacrifice themselves to ensure the reproduction of others, which are closely related and thus have a high chance of possessing and passing on the same genes (Velicer, 2005). In soil samples of Dictyostelium relatedness is 0.52, meaning that individuals on average share 52% of the same genes. The complex network of protein interactions which allows development to occur, however, also encourages fruiting bodies to comprise cells that are even more closely related. This significantly raises the average relatedness of cells within the same fruiting body (Gilbert et al., 2007). One interesting side effect is that if two different cultures of Dictyostelium are mixed prior to development, one of the cultures will preferentially make spore cells while the other will make stalk cells, even though normal distribution of stalk and spore cells occurs when they develop independently (MacWilliams et al., 2006) Dictyostelium discoideum Development Most of the timing of key events in the Dictyostelium discoideum developmental process is externally mediated as opposed to regulation by internal timers (Jain et al., 1992). It has been known for some time that purinergic signals play a vital role in this mediation. Although a wide range of chemoattractants are used to regulate development in slime moulds, the group 4 species (which include Dictyostelium discoideum) use cyclic AMP (camp) (Schaap, 2011). The emission of this camp is triggered by alterations in the ratio between the concentration of the constantly emitted autocrine Prestarvation Factor (PSF) and the concentration of prey bacteria. PSF levels are themselves proportional to the cell density and in general PSF/bacteria ratios reach the critical point three to four generations before the bacterial levels become too low to support further exponential Dictyostelium growth (Burdine and Clarke, 1995). Dictyostelium discoideum posseses four camp receptors known as car1-4, each with different affinities to the agonistic camp. For the initiation of development car1 and car3 are the most important with double mutants failing to chemotax or aggregate properly (Manahan et al., 2004). car1 is coupled to the g-grotein Gα2βγ on the inside of the membrane. Upon camp binding, the Gα2 subunit separates from the bulk of the protein (Ray et al., 2011), triggering the activation of adenyl cyclase (ACA) which generates further camp. This camp is then exported outside the cell where it is able to spread through the medium thus propogating the signal (Das et al., 2011). Simultaneously, the gene pdsa encodes a phosphodiesterase (PDE) to degrade the extracellular camp and ensure that later releases of the signal cause a detectable change in concentration (Bader et al., 2006). Another effect of camp binding is to trigger phosphorylation of the car1 receptor, lowering its affinity to camp until it is rectified. This causes the receptors responses to occur in bursts upon exposure as opposed to a constant release (Das et al., 2011). Bursts of camp, closely followed by camp degrading phosphodiesterases, result in the formation of the camp waves for which Dictyostelium is famous. Two splicing variants of the phosphodiesterase exist: a secreted one and a membrane bound one. Mathematical simulations suggest that the reason for these two forms is to provide optimal activity at different cell densities (Palsson, 2009). This expression of both adenyl cyclase and phosphodiesterase results in the creation of waves of camp spreading throughout the medium from multiple aggregation centres. Like all waves, when these meet they experience both constructive and destructive interference: areas of high and low camp concentration (Palsson, 2009). The cells are capable of detecting differences in concentration across their length, by determining how many of the Gα2 receptors are activated on each side (Kortholt et al., 2011). The cells are capable of 18

19 detecting very small differences in camp concentrations because the cells themselves are tiny compared to the wavelengths of the camp waves. They then chemotactically move in the direction of the higher camp concentration. This movement can only occur when the camp waves are passing over the cells and the car1 receptors are active. The result is periodic inward movement, which can be visualised using a timelapse microscope (Dormann et al., 2000). The movement of cells emitting camp changes the location of the high and low concentration areas, but the eventual result is the formations of clusters of up to 1x10 5 cells around aggregation centres. Careful observation of cells reveals that when the waves are passing over them and they are undergoing chemotaxis, they also undergo morphological changes. They became elongated and this changes their brightness when observed under a dark field microscope, allowing one to visualise the shape and progress of the camp waves (Dormann et al., 2000). As the elongated cells converge upon the aggregation centre they connect head-to-tail forming, what are known as, chemotactic streams that extend the entire length of the aggregation territory. These streams help guide the cells efficiently towards the centre (Veltman and van Haastert, 2008) and by continuing to emit camp they greatly increase their range. When the cells enter the streams they start differentiating into the prespore and prestalk cells, which will eventually further differentiate into the spore and stalk cells. There is no correlation between their tendency to form one type or the other and their position within these streams (Weijer, 2009). At the aggregation centre, where the streams meet, they form a hemispherical mass known as the mound (Siegert and Weijer, 1995). Cells within the streams are known to adhere to each other and this is mediated by multiple proteins, normally classifed by their resistance or lack of resistance to dissociation by EDTA. There is some evidence to suggest that EDTA-resistant proteins maintain end-to-end cell contact, whereas the EDTA-sensitive proteins are involved in mediating contact between side-to-side cells (Brar and Siu, 1993). Gp24 is one such adhesion protein, which also goes by the names of DdCAD-1 and contact site B (csb) (Coates and Harwood, 2001). It is encoded by the gene cada, which starts being expressed upon starvation in Dictyostelium grown on bacterial cell lawns. Expression of the mrnas peaks, and starts to fall again, three hours before protein levels peak, but this peak level stays near constant due to the protein having a suprisingly large half life (Yang et al., 1997). Curiously cada is also expressed during vegetative axenic growth at high cell densities, being induced by the PSF levels (Brar and Siu, 1993). Over the course of aggregation the cytoplasmic gp24 is relocated to the plasma membrane (Sesaki and Siu, 1996). Gp24 is EDTA sensitive, which means that it is Ca 2+ dependent. It contains two Ca 2+ binding domains (Sesaki and Siu, 1996). Experiments in which non-membrane-bound gp24 is added to cells show a large reduction in cell adhesion, implying that the adhesion is created by homophilic interaction with the same protein on the membrane of neighbouring cells. There is only one binding site per protein. In these experiments only small aggregates of between 3-10 cells formed and over half of the cells present failed to develop (Brar and Siu, 1993). Blocking of gp24s binding sites does not prevent the formation of aggregates, but they are far looser and less compact (Siu et al., 1988). The initial camp pulses also trigger expression of receptors, allowing the detection of DIF-1, which is needed to form two of the four subtypes of prestalk cells. Differentiation into prespore cells also requires extracellular camp, but there are several other compounds that are essential for it to occur (Yamada et al., 2010). 19

20 The mound forms a moving slug shape. The tip of the slug acts as a pacemaker, continuing to secrete camp periodically. The waves flow back along its length showing the later cells the direction in which to move (Jaiswal et al., 2012a). The slug moves to a point of high ground following light and temperature gradients (Flegel et al., 2011). As it moves the cells are sorted so that the front is composed mostly of the prestalk cells, while the posterior contains the prespore cells (Yamada et al., 2010). Once the slug has reached a suitable place to form a fruiting body it stops moving, settling on its posterior and forming the so-called mexican hat stage. The tip begins to secrete a cellulose tube and the cells begin to ascend. The prestalk cells which were the closest to the tube initially end up around the base, while the further away prespore cells have further to climb to find a space and end up at the top (Tyler, 2003). 1.3 There Are Potentially Further Purinergic Signalling Systems Regulating Development In Dictyostelium discoideum camp is not the only purinergic molecule known to have an effect on the development of Dictyostelium. Adenosine is known to promote the formation of larger aggregates as well as inhibiting the synthesis of the camp degrading phosphodiesterases (Jaiswal et al., 2012b). In 2007 bioinformatic techniques were used to detect five Dictyostelium discoideum genes that were very weakly homologous to human P2X receptors. When cloned in human HEK cells, these genes were confirmed to indeed be ATP receptors (Fountain et al., 2007). At the time it was believed that these receptors could be involved in cell-cell communication. The proteins encoded by these genes named p2xa, p2xb, p2xc, p2xd and p2xe, however, were found to all be located on the inside of the cell. They are now believed to be attached to internal vacuoles and control the release of Ca 2+ ions from acidic stores (Sivaramakrishnan and Fountain, 2012). The P2X receptors were initially believed to be involved in cell-cell communication because prior to their discovery there was already some evidence to suggest that exogenous ATP is capable of speeding up the onset of aggregation in Dictyostelium discoideum. Furthermore, the presence of ATP is known to cause an increase in the number of phosphorylated camp receptors earlier in development (Mato and Konijn, 1975). As well as this direct evidence for the role of extracellular ATP, it has been known for some time that Dictyostelium discoideum possesses cell surface ATPases (Parish and Weibel, 1980). It seems unlikely that Dictyostelium discoideum would possess proteins that have no use and this can be construed as secondary evidence for an extracellular ATP signal for these ATPases to terminate. Following the discovery of the intracellular location of the P2X receptors, experiments were carried out to confirm that these earlier findings had been correct. Attempts were also made to explain these results using advances in the field of molecular biology made since the original observations. It was found that upon exposure to ATP, cells would react with a large and rapid increase in intracellular Ca 2+ levels. At the same time it was discovered that ADP was capable of producing an almost identical calcium response while other nucleotides could not. The fact that the ATP response could be desensitized by exposure to ADP implied that both nucleotides were detected by the same receptor. This response was, however, inhibited by the addition of the P2X inhibitor gadolinium, implying a similiarity between the internal P2X receptors and whatever unknown receptor was giving this response. Further evidence that this response was initiated by a P2X receptor was provided by mutants which lacked the g-proteins required by P2Y receptors but still produced the calcium 20

21 response. It was, however, conclusively shown that the discovered P2X receptors were not involved in this response as mutants lacking these receptors still give it (Ludlow et al., 2009). The mechanism allowing Dictyostelium discoideum to undergo a P2X-like response without using P2X receptors is unknown. The evidence is mounting that Dictyostelium discoideum uses extracellular ATP for an unknown purpose. In this project an investigation was launched into firstly discovering the circumstances under which extracellular ATP was released, and secondly the effect of removal of this extracellular ATP. Attempts were then made to discover the mechanisms responsible for this effect and from this deduce the role of extracellular ATP under control conditions. 21

22 Chapter 2 - Materials And Methods Cell Culture A LB agar plate was completely spread with Klebsiella to act as a food source for Dictyostelium discoideum. Frozen stocks of the amoeba containing 1x10 7 cells were then thawed on ice. Individual drops of this solution were then added to five evenly spaced areas of the plate. This plate was left in the 22 O C incubator to allow the Dictyostelium to grow and spread across the whole surface. The rest of the thawed stock of cells were added to a petri dish already containing 10ml of HL5 growth media, 100U/ml penicillin and 100µg/ml streptomycin. The following day, when there were plenty of cells adhering to the bottom of the plate, the liquid from this media was added to a new plate and fresh media, penicillin and streptomycin added to the original plate. When spores had formed on the bacterial lawn plate, they were carefully removed and suspended in another plate also containing 10ml of HL5 media and 100U/ml penicillin and 100µg/ml streptomycin. The first of these three plates to produce a large culture of healthy looking cells (determined by microscopy) was then used for experiments. The soup was removed from the healthy cells and discarded to eliminate any lingering bacterial contaminants. The cells adhering to the plate were washed and poured into a conical flask containing 20ml of HL5 media along with 100U/ml penicillin and 100µg/ml streptomycin. A sponge bung was placed on the flask to prevent contaminants from entering, while allowing fresh O 2 to enter and CO 2 to leave. This flask was then left on a shaking incubator set to 150RPM at room temperature. After 6 hours the cell density was calculated with a haemocytometer and, if necessary, more media was added to the flask to adjust the cell density to 1x10 6 cells/ml. Once the culture was confirmed to consist of healthy cells, aliquots of 1x10 7 cells were taken and frozen in order to provide future cultures. Every 24 hours the cell density was calculated and cells were spun down in a centrifuge at 1500RPM for five minutes. The soup was removed and the cells were resuspended in a fresh 50ml flask of media with 100U/ml penicillin and 100µg/ml streptomycin. The number of cells spun down was calculated to be such that the cell density would be 1x10 6 cells/ml at the point when cells would be taken for experiments. Cells at a high density start to express developmental genes (Jain et al., 1992). Care was taken to ensure that cell density never exceeded 4x10 6 cells/ml as this was found to be the density at which this effect began to occur in these lab conditions. The 100U/ml penicillin and 100µg/ml streptomycin were added to the cell cultures to reduce the risk of them being infected with bacterial pathogens Use Of The Haemocytomer Before each use the haemocytometer was carefully cleaned and dried, then a clean coverslip was placed on top. 10µl of the sample was taken up with a pipette. The tip was carefully pressed against the side of the coverslip and the solution released, with care being taking to ensure the coverslip was not moved. For these experiments a two chambered haemocytometer was used, allowing two cell counts to be made before it needed to be washed and reloaded. Each side of the haemocytometer was loaded and then observed under a microscope. The grid upon the haemocytometer was located as in Figure 1a. 22

23 Figure 1 Use Of The Haemocytometer. a) The view of the counting grid of a haemocytometer loaded with Dictyostelium discoideum cells b) the same haemocytometer with the squares to be counted labelled The number of cells inside the four squares in the corners and the centre square were counted (Figure 1b). The top and left hand side of the squares were included while the other two sides were not. Each chamber was counted and then the haemocytometer was washed and reloaded so as to give a total of 4 counts. A mean of these counts was taken and then multiplied by 50,000 in order to give the number of cells found in 1ml of the sample Measuring Growth Curves Two 10ml conical flasks were each filled with 3ml of cells at 5x10 5 cells/ml. To one of the flasks was added 60µl of 200U/ml apyrase to give a concentration of 4U/ml. To the other flask was added 60µl of Molecular Grade water to act as a control. The number of cells in each flask was calculated at 10:00am and 16:00pm every day using the haemocytometer. For these experiments one monitored the growth of the same culture of cells in the same media and flask throughout. The cells cultured normally in the lab for use in experiments were transferred to fresh media and flasks daily to prevent the build-up of pathogens and to ensure they always had fresh nutrients in the media. This could not occur in this experiment as culturing would impact the growth curves and wash out the apyrase. Each experiment was therefore continued until the cultures in the flask were visibly unhealthy. This experiment was repeated four times. 23

24 2.3 - Development Buffer In many of the studies carried out, the cells were incubated in an isotonic development buffer. This consisted of Na 2 HPO 4 (5mM); KH 2 PO 4 (5mM); NH 4 Cl (18.75mM); MgSO 4 (328µM); CaCl 2 (252µM). This was adjusted to be ph6.8. This was then passed through a 50mm diameter filter with a pore size of 0.2µm in a UV sterilised hood into an autoclaved bottle. This bottle of buffer was kept at 0 O C until half an hour before an experiment, whereupon it was brought back to room temperature using a water bath. In early preparations of the buffer it was autoclaved instead of filtered, but this led to the formation of a cloudy precipitate, which interfered with development and when viewed under a microscope was visually similar enough to bacterial contamination to cause confusion in the interpretation of results. 2.4 Development In Submerged Culture Two and a half hours before the experiment began, the cells were cultured using the methods described to a density of 5x10 5 cell/ml. This was because pre-developmental cell density and the associated build up of PSF levels could influence the expression of certain developmental genes (Brar and Siu, 1993). By culturing the cells to the same density into fresh media a constant time before the experiments were due to start, one could attempt to minimise the effect of these factors and give consistency to repeated experiments. At the start of the experiment the cell density was again calculated and the required volume of cells centrifuged for four minutes at 1800RPM and then resuspended in 20ml of development buffer. This was pelleted again under the same conditions and the buffer discarded twice more. The cells were then resuspended in enough buffer to give a relevant volume of cells at an appropriate density for the particular experiment. The multiple centrifugations were to help remove any contaminants in the original cultures, as well as to wash off any trace volumes of the HL5 media as this contains ATP (Min et al., 2006), which could interfere with these experiments. A 24 well plate was filled with 1ml of the cell-buffer solution in each well the experiment required. The plate was then placed into an illuminated incubator set to 23 o C. Over time the cells settled to the bottom of the well and adhered to the base. Cells were regularly observed under a microscope and quickly returned to the incubator. Photos were taken of the cells under the microscope if relevant Treatment With Apyrase Apyrase was prepared at a stock concentration of 200U/ml in molecular grade water. This was then sterilised by passing the solution through a 0.2µm filter. The submerged culture experiments were set up as outlined in section 2.4. Half of the wells had 4U/ml apyrase added to them. These wells were labelled. The other half of the wells had an equal volume of molecular grade water added to act as a control. The Dictyostelium development within these wells was observed to see if the apyrase-treated cells displayed a phenotypic difference Treatment With Heat-Inactivated Apyrase Initial experiments were carried out in which apyrase was heat-inactivated at the stock concentration. This caused the protein to precipitate out of the solution as a sticky lump in which cells would get 24

25 trapped. In an attempt to counter this effect, for the heat-inactivation experiments, the apyrase stock was added to Eppendorf tubes of development buffer at a concentration of 8U/ml. These were placed in a 95 O C heatblock for ten minutes before being removed and allowed to cool slowly to room temperature. Wells were prepared as in section 2.1, but with the wells containing 0.5ml of cells at twice the density. 0.5ml of the heat-inactivated apyrase in solution was added to give a final density of both cells and apyrase as found in the previous experiments. For the sake of maintaining control conditions throughout the experiments, the well containg apyrasetreated cells were also prepared during these experiments at 8U/ml in development buffer with 0.5ml added to 0.5ml of cells at 1x10 6 cells/ml. The same was done for the control cells, but with unchanged development buffer added in place of apyrase solutions. 2.7 Development On Agar 1.5% mass per volume bacteriological agar was dissolved in development buffer. A microwave oven was used to heat the buffer sufficiently for the agar to dissolve. 3ml of this agar solution was added to each well of a 6 well plate and allowed to solidify for 1-3 hours. During this process the lid was placed back on the wells to stop them drying out. 2.5x10 6 cells were taken for each well used in the experiment and resuspended in development buffer at 1x10 7 cells/ml using the same washing procedure as in the submerged culture experiments. 0.2ml of this solution was then transferred to an Eppendorf for each well. If required, 4U of apyrase in 20µl in molecular grade water were added to each of these Eppendorfs. The contents of the Eppendorf were then mixed slightly with a pipette and added to the centre of one of the agar filled wells from earlier. A plastic spreader was used to ensure even distribution within the well. The plate was wrapped in damp paper to prevent the agar drying out, and placed in the incubator upside down so that evaporated moisture would condense near the cells. Development typically occurred over 2-4 days. The plates were observed and redampened frequently over this time Determining Fruiting Body Size Single fruiting body heads were extracted with a damp pipette tip and resuspended in 4ml of Klebsiella. 0.5ml of this solution was added to two agar plates and spread evenly with separate spreaders. These were left in an incubator until colonies developed. These colonies were counted and the mean number of colonies produced from each head calculated. Initially the heads were suspended in 1ml of Klebsiella which was completely spread on the plates, but with so much Dictyostelium on a plate, the limiting factor for growth became not the number of viable cells, but amount of space on the plate so a more suitable dilution was used. 25

26 2.9 Measuring Extracellular ATP ATP release was measured in submerged culture and exponential growth. In these experiments the cells were set up as in the relevant experiment above together with a control sample of just the relevant media or buffer. At each timepoint 50µl of the solution was taken from both the sample and the control and added to individual fresh Eppendorf tubes. These were then centrifuged at -4 O C for 4 minutes at 2300RPM in order to separate the cells and their contents from any secreted products. The low temperature was used in order to limit natural ATP degradation. Once the centrifugation was complete, 25µl of the supernatant was removed from the Eppendorfs and placed into a fresh Eppendorf. 25µl of 0.1mg/ml luciferase was added to each Eppendorf and they were placed in a biophotometer in order to measure light emitted with an integration time of 7 seconds Measuring Cell Adhesion Aggregation Quotient Aggregation was calculated according to an Aggregation Quotient derived by comparing the number of lone cells under the haemocytometer to the number of groups of cells. This was used instead of comparing the number of lone cells to the number of cells adhering to each other because it was hard to accurately determine the number of cells making up the larger groups. The Aggregation Quotient was calculated as follows: It is important to note that the Aggregation Quotient is the ratio of aggregates to lone cells, as opposed to a measure of how many cells have aggregated. An Aggregation Quotient of 1 would mean that all the cells present had become part of aggregates and one of 0 would mean that none had. A quotient of 0.5 would mean that there were an equal number of aggregates as lone cells as opposed to 50% of the total cells being aggregated, which one might expect it to mean in other systems. The Aggregation Quotient at time zero was never found to be zero. This was because some cells were in the process of splitting into two via binary fission. It is not believed that an increase in Aggregation Quotient could be caused by this binary fission process because this only occurs during the axenic growth phase and cannot occur in the buffer-induced starvation conditions (Saxer et al., 2010) Measuring Adhesion A relevant volume of cells were suspended at 5x10 5 cells/ml in development buffer. 1.5ml of this solution was then added to 2ml centrifuge tubes. Chemicals such as EDTA and apyrase were added at this stage as required. These tubes were attached to a vertical rotating wheel in such a way that each half turn would invert the tubes. The 0.5ml air pocket was to ensure that the contents would be adequately mixed. This device was set to 46 RPM. Every 30 minutes the device was stopped and 10µl of the tubes contents loaded onto a haemocytometer slide. The number of groups of cells and the number of lone cells were counted and the Aggregation Quotient calculated. Three tubes of each treatment were set up in each experiment, so as to minimise variation between tubes. 26

27 Measuring ATPase Activity Phosphate Liberation Buffer Phosphate liberation buffer consists of Tris-maleate (0.1M); MgCl 2 (2mM); KCl (2mM); ph 7.2 with KOH. The ph of the buffer was adjusted to 7.2 by careful addition of KOH. This buffer precipitates if left at room temperature for a couple of days, so it was made up either on the day of the experiment or the night before and kept on ice until needed MLG Reagent MLG Reagent consists of malachite green (1.0mM); polyvinyl alcohol (0.16% w/v); H 2 SO 4 (6.0mM) Malachite Green is a highly potent dye and stains everything it touches. It is imperative that gloves are worn during the preparation of this reagent and that all skin is covered Molybdate Solution Molybdate solution was prepared consisting of 50.0mM ammonium molybdate dissolved in 3.4M H 2 SO Measuring Phosphate Liberation A 24 well plate was prepared with 2 wells containing phosphate liberation buffer alone; one well containing cells at 1x10 6 cells/ml in phosphate liberation buffer and one well containing 2x10 6 cells/ml in phosphate liberation buffer. The plate was incubated at 23 O C for one and a half hours to allow the cells to adhere to the wells and to allow the cells to condition the buffer with ATP. The plate was then placed on a shaker at 45RPM. 20µl samples were taken from each of the four wells to count as time zero. One of the wells containing only buffer had its sample used as a blank to calibrate the spectrometer. The other was used as the experimental control well to measure phosphate liberation in the absence of cells. 0.5mM ATP dissolved in the phosphate liberation buffer was added to the three experimental wells and samples were taken every three minutes. After a sample was taken it was placed on ice for 15 minutes to limit spontaneous breakdown of ATP. After the incubation, the samples were removed from the ice and 2µl of Molybdate Solution were added. They were then incubated at room temperature for 10 minutes. Following this 3.71µl of MLG reagent was added and they were incubated at room temperature for another 15 minutes. Finally 20µl of the resulting solution was added to a cuvette and diluted with 480µl of distilled water. This was given a final incubation of five minutes. The absorbance of light at a wavelength 600nm was measured using the spectrometer. It was important that the timings were followed exactly with each sample, as absorbance was observed to increase the longer the samples were incubated, regardless of phosphate concentration Data Analysis In science one cannot do an experiment only once and draw conclusions as one cannot tell whether the achieved result arose due to chance. Each of the experiments observing a phenotype, carried out in 27

28 this project was carried out a minimum of three times in order to ascertain that the same result occurred each time, often many times more. The exception to this rule was the experiment involving the heat-inactivation of apyrase, which was limited due to the high rate of failure in preparing the protein for treatment. Some of the results of the experiments are more quantitative. Instead of observing describable differences in the cells the results are numeric, for example: the number of aggregates produced from the same volume of cells under different conditions. For these experiments the results have been given in the form of a graph of an appropriate type for the data. Unless otherwise stated, the graph will typically display the mean data for each treatment, with error bars showing the standard error of the mean of the data for that treatment. The standard error of the mean is calculated as the standard deviation of the data for that treatment divided by the square root of the number of data points generated for that treatment. The number of times the experiment was repeated to generate the data is recorded in the form n=the number of repeats in the caption for each graph. When numbers have been generated by the experiments instead of descriptive data, statistical analysis can be carried out to determine how likely it is that differences in the means of the data sets is due to an actual scientific difference as opposed to random chance. Unless otherwise stated, the statistical analysis on the data sets in the graphs is a Student s T-Test, paired or unpaired depending on whether the data was generated in paired sets. Statistical significance between data points is marked on the graphs by an asterisk (*). One asterisk represents a P value (the likelihood of the difference in means being a result of chance) of less than Two asterisks represents a P value of less than 0.01, and three represent a P value of less than

29 Chapter 3 Investigating The Presence Of Extracellular ATP And The Effect Of Its Removal Upon Growth And Development In Dictyostelium discoideum Aim Initially the aim of the experiments described within this chapter was to determine whether, and under what conditions, Dictyostelium discoideum cells secrete ATP. Following the discovery of this, the impact of the removal of this ATP was then investigated. 3.2 Introduction The purpose of the initial experiments in this chapter was to ascertain whether extracellular ATP had a role anywhere in the Dictyostelium discoideum life cycle. There is already evidence to suggest that extracellular ATP is released during axenic growth (Parish and Weibel, 1980) and the first experiment carried out was to confirm that this is the case. Experiments were also designed to ascertain whether extracellular ATP was released during the developmental phase of the life cycle. The presence of extracellular ATP in both instances was determined via the luciferase assay. Luciferase is an enzyme found in fireflies, which reacts with ATP to produce bioluminescence. The reaction is as follows: One of the more useful attributes of this reaction is that the amount of light produced is directly proportional to the amount of ATP present. Unknown concentrations of ATP can be added to solutions and the intensity of the light measured with a photometer (Turman and Mathews, 1996). By comparing this to a standard curve of the light produced by known concentrations of ATP in the same buffer one can deduce the concentration of ATP in said solution (Nichols et al., 1981). Once extracellular ATP was detected in a particular stage of the Dictyostelium life cycle, it was removed from the medium using the potato enzyme apyrase, and the cells observed to see if there was any change in their condition. Apyrase is an ATPase, which catalyses the breakdown of ATP into ADP and phosphate and then ADP into AMP and more phosphate (Riewe et al., 2008). The effect of removing extracellular ATP during the axenic growth phase was investigated by measuring the growth rate. Under normal circumstances cells from the same culture suspended in flasks kept in the same conditions will have the same doubling time. If degrading extracellular ATP were to cause any alterations to the cell cycle, then the timing of the cell cycle phases would differ and as a result the doubling time would change. By comparing the growth curves of normal cells and apyrase-treated cells one could ascertain whether the apyrase was having an effect. In order to investigate the developmental processes of Dictyostelium submerged culture was used. Submerged culture is a method where cells are suspended in a development buffer and added to wells of a plate. The cells sink to the bottom of the well, adhere to the base and start to form streams and aggregates. The buffer provides a good medium for the propagation of the camp signals while 29

30 Luminescane (RLUs) the clear buffer and plate allow for easy observation of stream formation and aggregates. The major downside of the use of submerged culture is that when cells are submerged in liquid, development does not typically occur beyond the aggregate stage (Sternfeld and Bonner, 1977) and therefore different techniques must be used to observe cells in the later stages of development. Subsequent to this experiment, investigation was made into the effect of apyrase on Dictyostelium growth on agar, firstly to see whether extracellular ATP played a different role depending upon whether the cells were submerged and secondly it was important to see whether this ATP played a role in the later stages of the developmental cycle: slug and fruiting body formation. 3.3 Dictyostelium discoideum Undergoing Exponential Growth Causes Luciferase To Produce More Luminescence Than Untreated Buffer Shaking culture flasks were set up containing 3ml FM minimal media and Dictyostelium discoideum cells seeded at an initial density of 5x10 5 cells/ml. These flasks were shaken at 150RPM for 17 hours. Samples were then taken from these flasks and from flasks of media that had undergone the same conditions without cells. The luciferase reaction was carried out upon these samples. A significantly higher luminescence from a sample would be indicative of significantly higher ATP concentrations within that flask. In this laboratory HL5 media was normally used for cell culture, but in these experiments FM minimal media was used instead. This was because HL5 is made from yeast extract which is known to have a high ATP content (Min et al., 2006). Use of HL5 would raise the background luminescence and drown out any contribution to extracellular ATP concentrations made by the Dictyostelium cells * Media Cells Figure 2 Increased Luminescence From The Luciferase Reaction In Culture Medium Conditioned by Growing Dictyostelium. The luciferase-luciferin reaction results in significantly less luminescence in unconditioned FM minimal growth media than media conditioned with growing Dictyostelium (5x10 5 cells/ml; 17 hours at 125RPM). The error bars represent standard error of the mean of each data point, N=4, P<

31 Cells/ml The samples from the flasks that had contained cells produced a significantly higher (P<0.05) amount of luminescence than the samples from the flasks without cells (Figure 2). It is important to note that this data does not provide the ATP concentration presents in the media either with or without cells. It merely shows that after 17 hours the buffer containing cells causes luciferase to produce significantly higher luminescence than buffer which does not have cells. This is, however, indicative of there being more ATP present in the cell-conditioned buffer than the un-conditioned buffer. 3.4 The Addition Of Exogenous Apyrase Has No Effect On Exponential Growth Of Dictyostelium discoideum Having found evidence to suggest that during growth in shaking culture the cells conditioned their media with ATP, the effect of removing this ATP was now investigated. This was done using 4U/ml apyrase. Cells were suspended in flasks of the FM minimal media at 5x10 5 cells/ml under control conditions or in the presence of 4U/ml Apyrase. The cell density was measured at regular intervals over a period of 54 hours. FM minimal media was again used instead of the HL5 for the same reasons as in the previous experiment Control 4U/ml Apyrase Time (Hours) Figure 3 Growth Curves Of Dictyostelium. Growth over a period of 54 hours in FM minimal media under control conditions or in the presence of 4U/ml apyrase. The error bars represent standard error of the mean of each data point, N=4, P>0.05. No significant difference was found in cell density, at any of the time points used, between cells grown under control conditions or with 4U/ml apyrase (Figure 3). Apyrase does not affect the growth rate of Dictyostelium within the first 52 hours following seeding of cell cultures. 31

32 Concentration of ATP released (nm) Dictyostelium discoideum Secretes ATP During Development Given the lack of response to removal of ATP during exponential growth the next stage of the Dictyostelium life cycle was investigated: development. Again the presence of ATP was investigated using the luciferase assay. A 24 well plate was set up with wells containing either 1ml of pure development buffer or 5X10 5 cells and was gently shaken to ensure that any extracellular products of the cells would have an equal concentration throughout the well. Samples were taken of the liquid within these wells every 15 minutes and their luminescence when mixed with luciferase was recorded. A standard curve of ATP in development buffer was also used to convert the luminometer s readings in RLU into actual ATP concentrations Time since addition of cells to buffer(minutes) Figure 4 - Detection Of ATP Upon Addition Of Dictyostelium To Development Buffer. Mean ATP detected by luciferase-luciferin reaction (luminescence, RLU) over the 120 minutes following seeding of Dictyostelium into development buffer at 5x10 5 cells/ml. Each data point is subtracted from the value obtained at the same time point from buffer that did not contain cells. The error bars represent standard error of the mean of each data point. Figure 4 demonstrates the change in ATP concentration recorded in the 1ml of development buffer containing 5X10 5 cells compared to wells of pure buffer. Immediately after addition of cells to the buffer, there is a very large change in ATP concentration which implies immediate ATP release. This sudden increase in ATP concentration, though consistently shown in each experiment was highly variable in magnitude. A student s T-Test revealed that over the entire two hour observation period, the ATP concentration was significantly higher than in buffer without cells (p<0.001). Repeated runs of the experiment all showed a high ATP concentration immediately after the cells entered the buffer. The ATP concentration then dropped and levelled off. A two-factor with replication ANOVA test revealed that over the entire two hour period the decrease in ATP concentration was statistically significant (p=0.01). On the other hand, if one only takes the data points between minutes into account the change over time was insignificant (p=0.96), implying that this fall in ATP concentration levelled off at the 60 minute mark. 32

33 Extracellular ATP in the presence of Dictyostelium discoideum is known to have a short half-life due to the presence of ecto-atpases (Parish and Weibel, 1980) so the fall in ATP concentration is not unexpected. The fact that the fall stopped, and the ATP concentration levelled off, suggests that the cells were releasing further ATP at a rate proportional to the rate of decay Submerged Culture Submerged Culture - Characterising The Control Before any effect of the addition of apyrase to Dictyostelium discoideum during development could be investigated, the correct experimental procedure as well as the appearance of development under control conditions had to be determined. Submerged culture was prepared in 24 well plates as described in the scientific literature at a density of 1x10 6 cells/ml. The experiment was set up in the morning and photos were taken over the course of that day and the following morning (Figure 5). Figure 5 Development In Submerged Culture. Dictyostelium development in submerged culture at cell density of 1x10 6 cells/ml. Photos taken at 4, 7, and 24 hours after addition to the wells. Images are annotated with descriptions of the developmental process. It quickly became apparent that, though aggregation occurred over this time period, it was at a rate that was unsuitable for observing the key events. As a result of this it was decided in the second experiment to use a variety of different cell densities as cell density is known to impact the rate of development (Jain et al., 1992). 33

34 Figure 6 - Developmental Progress Of Dictyostelium After 14 Hours At 4 Different Cell Densities (cells/ml). Aggregates of cells are visible at densities of 1x10 6 cells/ml and 5x10 5 cells/ml. On this occasion the cells developed faster than in the previous experiment. The cells that had been seeded at 1x10 6 cells/ml and 5x10 5 cells/ml had formed aggregates by the time they were observed at 14 hours (Figure 6), whereas the cells at higher densities were already too closely packed together for aggregation to be a meaningful term. Despite this, one could clearly see dark patches where the cells were at a higher density, implying that cells had moved closer together. Carrying out experiments at these higher densities, however, was impractical because these cells did not form noticeable streams. Attempts were made to perform the experiment at lower cell densities, but below 5x10 5 cells/ml the cells did not develop, even after several days of observation. It is assumed that this was because the cells were too far apart for camp waves to propagate. It was therefore decided to carry out the experiments at densities of both 5x10 5 cells/ml and 1x10 6 cells/ml because, although they had still developed at an impractical rate for observations of all processes to be made, it was hoped that the early stages of aggregation could be visible in 1x10 6 cells/ml and the later ones in the 5x10 5 cells/ml ones. Although the cells at 1x10 6 cells/ml again developed too quickly, this time development of the cells seeded at 5x10 5 cells/ml occurred slowly enough to be observed (Figure 7). 34

35 Figure 7 Development Of Dictyostelium In Submerged Culture At Cell Density Of 5x10 5 cells/ml. Between hours cells begin to elongate and lose their amoeboid shape. At 18 hours the elongate cells begin to sort themselves into patterns and very early streams, only a couple of cells long, can be seen. By 20 hours the streaming is occurring, the cells are forming a vortex shape as they determine where the eventual aggregation centres will be. These aggregation centres are visible at 22 hours and the cells are moving inwards along the streams towards them. By 24 hours the streams have begun to break up, each segment will form their own separate aggregate. Dictyostelium is notorious for being highly variable in the timing of developmental events due to minute changes in factors such as laboratory temperature, lighting or cell density of the culture from which they were taken (Jain et al., 1992). Great variation in the timing of development was observed across multiple experiments under what were, theoretically, the same experimental conditions. Investigation showed, however, that the cells in separate wells would start to stream within half an 35

36 hour of each other if the cells were taken from the same flask and added to the wells at the same time. Despite the fact that there was great variety in developmental timing between experiments, comparisons could still be made between cells in different wells if they were part of the same experiment Submerged Culture The Addition Of Exogenous Apyrase To Dictyostelium Cells Causes A Developmental Delay Submerged culture was set up at a density of 5x10 5 cells/ml with two wells containing cells treated with 4U/ml of apyrase dissolved in 20µl of molecular grade water and two wells of control cells containing an equal volume of molecular grade water with no apyrase. The development of all these cells over a period of fourteen to twenty-four hours was observed. Cells developing under control conditions formed significantly earlier than those treated with apyrase (Figure 8). Even though stream formation did eventually occur, the onset of it was inhibited in some way by the addition of apyrase. It was also observed that, while the cells in the control wells were all at the same stage of development, the apyrase-treated ones were far less uniform with completed aggregates sharing a well with early streams and cells that were still only elongating. 36

37 Figure 8 Apyrase Causes A Delay In The Formation Of Streams. Development of AX4 cells at a density of 5x10 5 cells/ml in cultures treated with 4U/ml of apyrase or an equivalent volume of water over a period of 24 hours. Time since addition to the buffer is given in hours on the left of the image. Photos are of the most developmentally advanced cells found within the well. In control wells streams are first visible at 20 hours and begin to break up by 24 hours. The Apyrase-treated cells, however, take a further two hours before they start to form streams. 37

38 It is important to note that if left long enough, the apyrase-treated cells eventually formed aggregates phenotypically similar to the control ones (Figure 9). Figure 9 No Effect Of Apyrase On Complete Aggregate Formation. After 30 hours in development buffer, aggregates of control Dictyostelium cells at a density of 5x10 5 cells/ml do not have observable differences to those formed in the presence of 4U/ml apyrase The most striking effect of adding apyrase to cells was apparent between one to three hours after addition to the buffer. While the control cells did not appear to have changed visibly, the apyrasetreated cells came together in a phenotype that was referred to as clumping (Figure 10). Figure 10 Apyrase Causes The Formation Of Clumps. After treatment with 4U/ml apyrase, Dictyostelium at a density of 5x10 5 cells/ml form "Clumps." Images taken 1.5 hours after additon of cells to the buffer Although hard to quantify scientifically, the wells with apyrase added to them appeared to contain a larger number of undeveloped cells following the formation of aggregates (Figure 11). 38

39 Aggregates formed Figure 11 - Evidence For Increased Incidence Of Cells Failing To Commit To Aggregate Formation Upon Treatment With 4U/Ml Apyrase. Cells at 5x10 5 cells/ml. Images taken at 20hours after addition to the buffer. It was assumed that a larger number of undeveloped cells would lead to the formation of fewer aggregates as this would lead to a functionally lower cell count during development. In order to determine if this was the case the number of aggregates formed in each treatment after 38 hours in development buffer across seven experiments were compared. There was a large amount of variation in the amount of aggregates formed in each experiment, presumably as a result of the same factors that caused the variation in the timing of stream formation between experiments. In order to be able to combine the data from multiple experiments, each data point was expressed as a percentage of the mean number of aggregates formed in that experiment s control wells. For this data analysis a total of 16 control wells at the 1x10 6 cells/ml density and 17 wells at the 5x10 5 cells/ml density were used. Each was paired with an apyrase well from the same culture of cells and under the exact same experimental conditions (Figure 12) Control Apyrase X105cells/ml 5 cells/ml Cell Density 1X106cells/ml 6 cells/ml Figure 12 Apyrase Induced Developmental Delays Caused No Significant Difference In The Eventual Number Of Aggregates Formed. Number of aggregates present in wells containing control and apyrase (4U/ml) treated cells following completion of development at two different densities of Dictyostelium. The error bars represent standard error of the mean of each data point, N=7. Each data point is expressed as a percentage of the average amount of aggregates within each pair s control well. 39

40 A two-tailed Student s T-Test of this data set confirmed that there was no significant difference between the number of aggregates found in the wells containing control and apyrase-treated cells with P values above 0.05 at both densities (0.61 and 0.34 at 1x10 6 cells/ml 5x10 5 cells/ml respectively) Submerged Culture The Addition Of Exogenous Apyrase Impacts The Development Of Dictyostelium Cells That Are Already Undergoing Development. When 4U/ml apyrase was added to Dictyostelium discoideum cells prior to the start of development, they experienced a delay in the formation of streams. This then posed the question whether apyrase would cause the breakup of already formed streams or whether it would only had an effect in the initial signalling phases prior to stream formation. In order to investigate this, 4U/ml of apyrase were added to wells that already contained streams. Large streams forming throughout the well, before the addition of apyrase Within an hour the streams have completely broken up After the break-up of streams, the cells start to form clumps Figure 13 Streaming Is Halted By Addition Of Apyrase. Apyrase (4U/ml) was added to Dictyostelium cells at a density of 5x10 5 cell/ml half an hour after the starting of stream formation. Time in hours since addition of cells to the buffer is recorded in the top left corner of each image. The steams very quickly broke up and the clumping phenotype was observed shortly afterwards (Figure 13). This indicates that apyrase had its effect on streams regardless of how far development had progressed at the time of addition. It is important to note that given enough time these cells also eventually formed complete aggregates Submerged Culture Heat Inactivated Apyrase Is Incapable Of Causing The Apyrase Induced Delay Having confirmed that adding apyrase to the wells caused a delay in the formation of streams of Dictyostelium discoideum cells it was now important to determine whether this was due to apyrase s enzymatic effect or whether it was due to the cells response to a foreign protein. Three sets of wells were set up containing firstly cells under control conditions, secondly cells with 4U/ml of apyrase and thirdly cells with 4U/ml of apyrase that had been heat shocked in order to denature it. 40

41 The cells that had been treated with heat inactivated apyrase developed at the same time as the control cells (Figure 14), whereas the undenatured apyrase-treated ones still suffered a delay. Figure 14 No Observed Effect Of Heat-Inactivated Apyrase On Development. Cells after 24 hours at 5x10 5 cells/ml under control conditions or with apyrase (4U/ml) that was either enzymatically active or heat inactivated This experiment proved harder to conduct than expected. When the heat inactivated apyrase returned to room temperature it would coagulate and form a sticky mess, trapping cells and preventing their development. This situation could be slightly improved by lowering the stock concentration of apyrase, but even this was not perfect and many experiments were ruined by this effect. As a result of this, and the relative cost of apyrase, only an n number of two was achievable with non-coagulated, heat treated protein Submerged Culture The Addition of Extracellular Adenosine Deaminase To Dictyostelium Does Not Counteract the Apyrase Induced Delay In Stream Formation. Apyrase breaks down ATP and ADP into AMP (Fenckova et al., 2011). Though apyrase was shown to cause a delay in stream formation, the question that remained was whether this was truly an effect of removing these nucleotides from the medium or whether it was a result of an increased build-up of the products of ATP hydrolysis. There is no known AMP receptor, but mammalian purinergic systems feature CD73 proteins which convert AMP into adenosine (Colgan et al., 2013). Adenosine is a purinergic signalling molecule with its own receptors (Burnstock and Verkhratsky, 2009) and is known to have developmental effects on Dictyostelium discoideum (Jaiswal et al., 2012b). It remained plausible that the apyrase increased AMP production, which in turn led to increased adenosine production, triggering adenosine receptors, which were then responsible for the delay. Whether increased adenosine production from increased AMP production was responsible for the phenotypes seen was tested by means of adding adenosine deaminase, which breaks down adenosine into inosine (Saboury et al., 2002), to the cells. Four types of well were set up containing cells under control conditions, cells with 4U/ml apyrase, cells with 2U/ml adenosine deaminase or cells with both 4U/ml apyrase and 2U/ml adenosine deaminase. The wells containing only adenosine deaminase were an important control to use. Otherwise, it could be suggested that any phenotype observed in the wells treated with both proteins was a result of extra inosine produced by adenosine deaminase, as inosine is known to affect Dictyostelium, though not during development (Parish, 1977). 41

42 Figure 15 Adenosine Deaminase Has No Effect On The Timing Of Dictyostelium Stream Formation. Photo taken 13 hours after addition to development buffer. In the wells that do not contain apyrase the cells are forming streams. In the wells that do contain apyrase there are no streams present but cells are forming clumps. Cells at a density of 5x10 5 cells/ml under control conditions or with apyrase (4U/ml) and with or without adenosine deaminase (2U/ml).. While the control cells and those treated with adenosine deaminase alone began streaming at the same time, all the apyrase-treated cells suffered a delay (Figure 15). This implies that break down of the adenosine building up in the wells of apyrase-treated cells had no impact on the timing of stream formation. Clump formation was also clearly visible in the wells containing cells treated with both proteins. 42

43 Aggregates Formed When aggregation was complete the number of aggregates present in the wells was counted. Curiously, although neither apyrase nor adenosine deaminase on their own caused a change in the number of eventual aggregates formed, when both proteins were added to the cells a significant (p<0.01) drop in the number of aggregates was observed (Figure 16) ** 20 0 Control Apyrase Adenosine Deaminase Apyrase + Adenosine Deaminase Figure 16 Additive Effect Of Apyrase And Adenosine Deaminase On Number Of Aggregates Formed During Development. Number of aggregates present in wells seeded with 5x10 5 cells/ml Dictyostelium under control conditions or treated with apyrase (4U/ml), adenosine deaminase (2U/ml) or both proteins after a period of 42 hours. The error bars represent standard error of the mean of each data point, N=3 with two wells in each experiment. The decrease in the number of aggregates produced by the cells that were treated with both proteins was not due to more cells making up each individual aggregate, indeed by eye the aggregates appeared to be a similar size to control aggregates, but was instead a result of the vast number of cells that had not taken part in aggregation. Whether this was due to a failure to express developmental pathways or due to cell death was unclear Submerged Culture The Effect Of Exogenous Apyrase On Dictyostelium Development Is Not Limited To The AX4 Strain So far these experiments had been carried out purely using the AX4 strain of Dictyostelium discoideum. There is, however, evidence that the AX2 strain releases higher concentrations of extracellular ATP during axenic growth than the AX3 strain (Parish and Weibel, 1980) and to have a significantly smaller calcium response to internal ATP than the AX4 strain (Sriskanthadevan et al., 2011). It was, therefore, wondered whether the apyrase effect was also found within AX2, or whether it was unique to AX4. The initial submerged culture development with apyrase was carried out again using AX2 cells (Figure 17). 43

44 Figure 17 - Apyrase Causes A Delay In The Initiation Of Streaming Of AX2 Cells. Development of AX2 cells seeded at a density of 5x10 5 cells/ml in cultures treated with 4U/ml of apyrase or an equivalent volume of water over a period of 24 hours. Time since seeding is given in hours on the left of the image. Photos are of the most developmentally advanced cells found within the well. Treatment with 4U/ml of apyrase had the same effect on both strains: the formation of clumps, followed by a delay in the initiation of streaming and the eventual formation of aggregates. It did appear, by eye, that the AX2 cells were forming larger clumps than the AX4 cells, but this is subjective. Furthermore, the large amount of variability in clump size between AX4 experiments makes it hard to determine whether this was an actual phenotypic difference between the two strains. 44

45 3.7.0 Cell Adhesion Cell-cell adhesion is an important element of Dictyostelium development. Cells start to adhere to each other when they first touch within the chemotactic streams and this helps to keep them travelling inwards towards the aggregation centre as well as to form fruiting bodies from the resultant mounds (Noratel et al., 2012). One suggested explanation for the clumping phenotype seen in cells treated with apyrase was that premature expression of the cell-cell cohesion genes was causing them to stick together during their random movement through the well prior to development. In order to test this an experiment was set up to measure the cohesiveness of cells when they collide during non-developmental conditions (Yang et al., 1997). It was hoped this would provide a method of quantifying the clumping effect and one could then use this to compare the degree of clumping between the AX2 and AX4 cell lines Cell Adhesion Characterising The Control Initially Eppendorf tubes were prepared containing 5x10 5 cells in 1ml of buffer to mimic the conditions found within the submerged culture experiments. It was found, however, that at this volume when the cells were inverted there were parts of the side of the tube that were never in contact with the liquid. The clear sides of the tube in these section became cloudy as the experiment continued and it was found that cells were entering these sections and not being resubmerged in the buffer. It was, therefore, decided to increase the volume to 1.5ml in order to ensure that every part of the tube s sides would be submerged with each full rotation. The number of cells and amounts of reagents were therefore scaled up to match this change in volume. As explained in section Aggregation Quotient, for these experiments the data for this section were measured as an Aggregation Quotient. This was a ratio of the number of aggregates seen within a sample of the Eppendorf s volume compared to the combined number of aggregates and lone cells seen. An aggregation quotient of 1 would mean that only aggregates were observed, while a value of 0 would mean that only lone cells were observed. Initally the experiment was carried out with the cells rotated at 10RPM, but no noticeable change in aggregation quotient occurred under these conditions. Yang et al. used 180RPM for their experiments, but here 46RPM was used. This speed was chosen as it was found to be the highest at which the tubes could be rotated before the centripetal forces pushed the contents of the tubes constantly to the ends and gave a centrifuge-like effect as opposed to a mixing one. 45

46 Figure 18 Aggregation Quotient Changes Over Time Under Control Conditions. Dictyostelium cells at a density of 5x10 5 cells/ml adhere to each other when mixed in the buffer through chance collisions and the expression of adhesion molecules a) Mean Aggregation Quotient over a period of 4.5 hours. The error bars represent standard error of the mean of each data point, N=8. b) Representative figure of the change of aggregation quotient in one tube over a period of 4.5 hours Figure 18 shows a sharp increase in aggregation quotient immediately after the cells are mixed in the buffer. This normally peaked between 1.5 to 2.5 hours and dropped slightly for another hour before levelling off. Experiments where the tubes were left on the rotor for over 24 hours before being monitored again suggested that the aggregation quotient remained at this level. It is important to note that this adhesion in the buffer is independent of camp waves as the constant mixing prevents the necessary camp gradient from forming. 46

47 Percentage of Control Aggregation Quotient at 3.5 hours Cell Adhesion Addition Of Exogenous Apyrase To Dictyostelium Does Not Alter the Aggregation Quotient Cell adhesion was measured over a period of 4 hours using cells under the control conditions described above paired with cells in the presence of 4U/ml apyrase. Each experiment included three tubes of each treatment Control Apyrase EDTA Time (hours) Figure 19 Addition Of Apyrase Has No Effect Upon The Aggregation Quotient. The change in Aggregation Quotient over time of Dictyostelium at a density of 5x10 5 cells/ml is not altered when apyrase (4U/ml) is added. The Aggregation Quotient is, however, altered by addition of EDTA (5mM) which prevents cell-cell adhesion. Control N=15, apyrase N=9, EDTA N=6. The error bars represent standard error of the mean of each data point. Due to variability in the size and the timing of the peaks in each experiment, each data point was expressed as a percentage of the mean control aggregation quotient at 3.5 hours within the current set of paired experiments. Figure 19 shows that no significant difference was observed between those tubes containing control cells and those containing apyrase-treated cells (p=0.62). In control experiments addition of 5mM EDTA caused a significant reduction in aggregation quotient versus untreated cells (p<0.001). EDTA has previously been shown to disrupt cell adhesion in Dictyostelium mediated by the gp24 adhesion molecules which is dependent upon divalent cations (Yang et al., 1997). This acted as a positive control to show that this assay was capable of detecting changes in the adhesiveness of the cells. With this in mind, it was speculated that cells treated with EDTA and apyrase would still form clumps in submerged culture, which would provide valuable evidence that the clumping effect was not due to the role of EDTA sensitive adhesion molecules. Development was observed in submerged culture at a density of 5x10 5 cells/ml, under either control conditions or with the addition of 5mM EDTA. 47

48 Figure 20 - EDTA Prevents Development. Dictyostelium seeded at 5x10 5 cells in submerged culture over a period of 20 hours either with or without exogenous EDTA (5mM). Control cells start to form streams at 15 hours and by 20 hours are beginning to become complete aggregates. EDTA treated cells fail to show any signs of development. The cells treated with 5mM EDTA failed to carry out any of the visible stages of the developmental cycle (Figure 20). Cells treated with EDTA did, however, differ from both control cells and cells treated with apyrase in that shortly after addition to the media they appeared to become more rounded 48

49 (Figure 21). This occurred even before the control cells had started to elongate and still retained their normal amoeboid shape. This rounded shape was retained throughout the time that control cells developed. It was, therefore, decided that the developmental phenotype of EDTA-treated cells was sufficiently different from control ones that any experiment treating cells with both apyrase and EDTA would give results more characteristic of the EDTA effect than that of apyrase. Figure 21 Rounded Appearance of EDTA Treated Cells. Within 1.5 hours of being added to the submerged culture, Dictyostelium seeded at 5x10 5 cells/ml and treated with EDTA (5mM) are noticeably rounder than ones in control conditions 49

50 3.8.0 Development On Agar Thus far the experiments had looked at the effect of apyrase on Dictyostelium discoideum development in submerged culture. This is not, however, the only method of investigating developing cells and in order to ascertain whether apyrase had an effect on the later stages of development the cells had to be observed when developing on agar Development On Agar The Addition Of Exogenous Apyrase Induces A Developmental Delay In Dictyostelium discoideum Developing On Agar Observing development on agar required larger wells than submerged culture (5cm diameter as opposed to 1cm). Although the cell surface density was kept constant, given that the purpose of this experiment was to develop cells that were not submerged in buffer, they had to be administered to the well at a higher concentration and lower volume (2.5x10 6 cells in 0.2ml of buffer). It was, at this stage, uncertain what volume of apyrase to add to match this alteration in well size. It was decided to initially start with 4U/well and alter it if this proved to not have an effect. Figure 22 Apyrase Causes Delayed Development On Solid Agar. Development of AX4 cells seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar. The cells treated with 4U/ml water are still at the mound and slug stages of development after 22 hours. At this time the control cells are mostly at the fruiting body stage with some remaining slugs. This was not necessary, however, as 4U/well was capable of eliciting a delay in development on agar, just as in submerged culture (Figure 22). This effect was most visible after 22 hours, when the control 50

51 Fruiting bodies formed cells had formed fruiting bodies with a few lingering slugs, while the apyrase-treated ones still mostly consisted of slugs with a few lingering mounds. Due to the three-dimensional structure of the fruiting bodies it proved difficult to focus on them to take photographs. In one of the control wells a fruiting body happened to form on the side of the well and, although the sideways gravity would affect growth of this fruiting body and it should be discounted as experimental evidence of a delay, a photograph of it is included in the control section of Figure 22 as representative of what was present in the well, but hard to photograph Development On Agar The Addition of Exogenous Apyrase Decreases the Number of Fruiting Bodies formed by Dictyostelium discoideum After five days of development, the number of fruiting bodies formed under control conditions and in the presence of apyrase was ascertained. The large size of the wells compared to the field of view under the microscope could render this a subjective measure. To counter this, the fruiting bodies were counted by volunteers who were uncertain which wells contained control cells and which contained cells treated with apyrase. As in previous experiments, due to the large amount of variation between different experiments each data point was expressed as a percentage of the number of fruiting bodies produced in the paired control * Control Apyrase Figure 23 Apyrase Causes A Decrease In The Number Of Fruiting Bodies Formed. Mean number of fruiting bodies formed by Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar with or without apyrase (4U/well). Each data point is expressed as a percentage of the relevant control. The error bar represent standard error of the mean the data, N=6, two wells/experiment. The cells treated with apyrase produced significantly fewer fruiting bodies (p<0.05) than the cells which developed in control conditions (Figure 23) Development On Agar The Addition Of Exogenous Apyrase Decreases the Number Of Viable Cells Within Fruiting Bodies Produced By Dictyostelium discoideum The mean size of the fruiting bodies formed by control and apyrase-treated cells was ascertained by taking two fruiting bodies of each treatment and re-suspending them in 4ml of Klebsiella. This solution was then spread onto agar plates and the colonies that eventually formed were counted. As in 51

52 Colonies formed previous experiments, due to the large amount of variation between different experiments each data point was expressed as a percentage of the number of fruiting bodies produced in the paired control *** 20 0 Control Apyrase Figure 24 - Evidence A For Smaller Number Of Viable Cells In Fruiting Bodies Developed During Apyrase Treatment. Number of colonies produced by 0.5ml of the mixture of a fruiting body head and 4ml of Klebsiella. Spores were produced from by Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar under control conditions or in the presence of 4U/well of apyrase. Each data point is expressed as a percentage of the relevant control. The error bars represent standard error of the mean the data, N=24 plates of control and 28 of apyrase. The fruiting bodies formed from cells treated with apyrase contained significantly fewer viable cells (p<0.05) than those formed under control conditions (Figure 24). 52

53 Colonies produced from a single fruiting body Development On Agar Investigating Whether There Is A Link Between Number Of Fruiting Bodies And Number Of cells Within The Fruiting Body Heads Given that the raw data was easily available from the previous two experiments, it was decided to investigate if there was a link between the number of fruiting bodies produced in a well and the number of colony forming units within each fruiting body head. Given that every well initially contained 2.5x10 6 cells, it seemed possible that the wells with fewer fruiting bodies would have more cells making up each fruiting body Apyrase Control Number of fruiting bodies produced from 2.5x10 6 cells Figure 25 - The Number Of Fruiting Bodies In A Well Containing 2.5x10 6 Cells Plotted Against The Mean Number Of Colonies Produced By Said Fruiting Bodies. Due to alterations in the set up for the colony counting experiment, only data from the final three of the six agar development experiments were considered. This still provided data on the number of fruiting bodies in 12 wells, as well as the number of colonies produced by four fruiting body heads from each well. The mean number of colonies/fruiting body was plotted against the number of fruiting bodies present in each well (Figure 25). Considering all twelve data points together there does not appear to be any link between the two sets of data. A Spearman s Correlation coefficient of 0.28 confirms a lack of correlation. When the data is split, to consider data from the control wells and the apyrase wells separately, there is still no link. These correlation coefficients are and respectively. 3.9 Conclusions It is believed that there is a role for extracellular ATP in the regulation of developmental timing in Dictyostelium discoideum. The initial evidence for this was in work carried out by Mato and Konijn in 1975 which showed that adding ATP exogenously causes an increase in the number of phosphorylated CAMP receptors. Furthermore, despite the absence of any known extracellular purinergic receptors in this organism, P2X and P2Y-like calcium responses to extracellular ATP were observed by Ludlow et al, in In this chapter experiments were carried out to investigate the role of extracellular ATP during development. The results from the luciferase assay suggest that the cells condition their media with ATP while undergoing normal growth. The addition of apyrase to the cells had no effect on the cells growth 53

54 rates. It should, however, be noted that a positive control showing that apyrase was capable of carrying out its enzymatic effect within HL5 media was not carried out. With this in mind it is not possible to conclusively say that breakdown of ATP was occurring under these conditions, one can therefore only definitively state that these results indicate that apyrase has no effect. If apyrase could indeed be shown to be breaking down ATP under these experimental conditions, with there still being no effect on the growth rate, then one must bear in mind that this experiment would only detect changes in growth rate within the first 52 hours of growth and between the densities of 5x10 5 and 3.5x10 6 cells/ml. Experiments could not be carried out over longer time periods, which would have allowed the cells to reach higher densities, because the cell cultures became less healthy over time. This was due either to consumption of all the nutrients within the flask of media or to the build-up of bacterial and fungal contaminants. Both of these problems are avoided during normal lab growth by means of regular culturing, but for the purposes of these experiments the cells had to remain within the same flask and volume of media for the duration. This, therefore, provides one possible explanation as to why addition of apyrase caused no observable effect. It might be that the extracellular ATP plays a role later in growth, when the cells reach a higher density. Indeed it is known that at higher densities, the growth rate decreases and the cells begin expressing a different set of genes (Jain et al., 1992). If detection of the high levels of ATP is responsible for this, then no evidence would be detected here when the cells were at lower densities. It is also important to note that the conditions in which Dictyostelium are grown in the laboratory are not the same as the natural conditions in which they evolved. In the wild the environment in which the cells grow is the same as the one in which they eventually develop. The luciferase assay also showed that the cells condition development buffer with ATP and that addition of apyrase at this stage has an effect. It seems plausible that the cells condition their environment with ATP during their growth phase in preparation for the role that it appears to have during development. These experiments suggest the existence of an ATP signalling system involved in the development of Dictyostelium discoideum. Upon entering starvation conditions the cells release, and more importantly maintain the level of, ATP in the medium. Although the ATP concentration measured in these experiments is very small, with a concentration measured in nanomolar, it is known that, in rats at least, prolonged exposure to similarly low levels of ATP will result in the activation of P2X receptors (Roberts et al., 2006). It therefore remains plausible that Dictyostelium could detect and respond to the low levels of extracellular ATP that the cells appear to maintain. The Apyrase used in these experiments breaks down both ATP and ADP at equal rates. These experiments cannot determine whether the effects of apyrase addition are a result of ATP breakdown or ADP breakdown. The luciferase assay suggests the presence of extracellular ATP in development buffer for apyrase to act upon, but extracellular ADP levels were not measured at any point during these experiments. Dictyostelium is known to have ecto-atpase activity but the same effect does not seem to break down ADP (Parish and Weibel, 1980) which would imply the presence of extracellular ADP under control conditions. Dictyostelium s calcium response to ATP can be induced equally well by ADP (Ludlow et al., 2009) which implies the two nucleotides activate the same unknown extracellular receptors. Without carrying out experiments in submerged culture using forms of apyrase that are preferential to one nucleotide over the other, one cannot conclusively say whether they both play the same role in these developmental experiments or whether it is removal of one in particular that is responsible for the observed apyrase effects. Addition of apyrase to submerged culture causes two developmental phenotypes distinct from that of the control: a delay in the formation of streams and the creation of clumps. It remains to be seen whether these two effects of apyrase addition are both a result of the same molecular pathway or instead represent interference with two different ATP/ADP mediated systems. Attempts to heatinactivate the protein suggest that this makes it incapable of causing the delay in stream formation, further suggesting that these phenotypes are caused by its enzymatic activity rather than the cells responding to the foreign protein. The fact this experiment was only repeated once means that this 54

55 result cannot be considered conclusive. Furthermore, the experiments in this chapter suggest that the phenotypes are not caused by the creation of excess adenosine formed from the excess AMP that is the end product of apyrase activity. Removal of adenosine by adenosine deaminase does not restore the control phenotype. One has reason to believe that this adenosine deaminase is indeed enzymatically functional, despite the lack of a positive control, as when it is exogenously added in the presence of apyrase one gets a lower final number of aggregates than otherwise. There could, however, be products of apyrase other than adenosine which could cause these developmental effects. There are no known AMP receptors (Colgan et al., 2013), but that is not evidence that they do not exist. An as yet undiscovered AMP receptor could be responsible for the apyrase phenotype. The other by-product of apyrase activity is phosphate ions. No reference in the literature could be found to the role of extracellular phosphate ions in Dictyostelium discoideum development. Phosphate-containing solutions are included, however, within the development buffer and it seems reasonable to suggest that alterations in the phosphate concentration could be responsible for the developmental delay. The experiment where adenosine deaminase is added as well as apyrase suggests that this explanation is unlikely to be the correct one. One knows that the adenosine deaminase is having an effect due to the alterations in aggregate number. If the apyrase effect was a result of increased phosphate ion generation then one would expect adenosine deaminase, which also produces phosphate as a by-product, to make the apyrase effect more pronounced. No difference in developmental timing between the apyrase-treated cells and the cells treated with both proteins was observed. Apyrase must be stored at low temperatures and denatures at room temperature (Aldrich, 2009). In these experiments the cells were kept at a constant temperature of 23 O C within the incubator. It is, therefore, possible that removal of ATP/ADP from the media completely prevents stream formation. Under this hypothesis the eventual stream formation and aggregation found in apyrase-treated cells is not symptomatic of a delay being overcome, but rather the cessation of nucleotide breakdown due to the apyrase becoming denatured. The experiments with heat inactivated apyrase suggest that the enzymatically inactive protein is incapable of delaying stream formation. Evidence to support this hypothesis can also be found in the fact that when apyrase is added to cells that are already streaming, the streaming stops. Alternatively this cessation of streaming in cells which have ATP and ADP suddenly removed may be caused by an abrupt change in gene expression. The hypothesis that the eventual stream formation is due to the apyrase becoming denatured could be tested by taking media samples for the luciferase assay from cells treated with and without apyrase throughout the developmental process. Regardless of whether apyrase becomes denatured during the developmental experiment or not, these results show that the pathway being investigated is important not just for the initiation of streaming, but in maintaining it as well. Shutdown of this pathway causes a rapid halt to streaming. These experiments suggest that ATP or ADP is needed for aggregation to occur at a normal rate. Furthermore they imply that either ATP/ADP or Adenosine need to be present for healthy aggregates to form. This can be demonstrated by the fact that while apyrase-treated cells had a delay in stream formation and adenosine deaminase treated cells had no observed phenotype, when both proteins were added to the cells there was a large drop in the number of aggregates they were capable of forming. This is curious because a cellular function which requires ATP/ADP or adenosine would be expected to indicate a pathway where ATP is broken down to adenosine. Addition of either purinergic molecule would have the same effect with the only difference being the point in the pathway at which this effect was induced. That the drop in aggregate number is not observed when adenosine is artificially broken down while ATP and ADP are not indicates that this is not the case. This implies the existence of redundant mechanisms to insure that whichever pathway ATP/ADP and adenosine use to regulate aggregate number under control conditions continues to operate in the absence of the detection machinery for either, but not both, of these molecules. This in turn suggests that this pathway is important as Dictyostelium has a relatively small genome and there is not much room for superfluous genes. 55

56 Further evidence to suggest that this difference in aggregate number is a result of interfering with purinergic signalling can be found in the fact that when cells are exposed to a higher concentration of adenosine than is natural, they form a greater number of aggregates which consist of fewer cells. It is also known that when this effect occurs, there is a greater number of solitary, unaggregated cells (Jaiswal et al., 2012b). It seems likely that this effect and the effect seen in this thesis, when adenosine, ATP and ADP are removed, are related. Despite the different developmental conditions, apyrase was able to cause a delay in development both in submerged culture and on agar. It is important to note that, while development on submerged culture showed no effects other than the delay, Dictyostelium development on agar produced significantly fewer fruiting bodies when treated with apyrase. This could be explained by either a fundamental difference in how development occurs in the two forms of culture or it could indicate that the apyrase-treated cells have difficulties in moving from the aggregate stage of development to the fruiting body stage. Testing this second hypothesis proved difficult as aggregates are not visible during development on agar. The number of mounds formed could not be used as a reliable method of determining aggregate number, as the number of mounds present at a given moment varied considerably over the time period of one experiment. It is known that there is a certain minimum size a fruiting body must attain, otherwise it collapses (Jang and Gomer, 2008). When the fruiting bodies were spread on bacterial plates, the ones formed by apyrase-treated cells only produced 40% of the number of colonies produced by control ones. This indicates that there are fewer viable spore cells within the apyrase-treated fruiting body heads, though whether this represents fewer cells overall or merely an increase in non-viable cells or stalk cells could not be determined. The fact that there fewer fruiting bodies formed when treated with apyrase and the observation that apyrase seemed to cause fewer cells seemed to join aggregates in submerged culture both provide valuable indirect evidence to suggest that apyrase decreases the number of cells that commit to aggregation. It was hoped, given that the number of cells present in each well remained consistent, there would be a correlation between the number of fruiting bodies present and the number of cells within them, and that apyrase would impact said correlation. There did not, however, seem to be such a correlation, perhaps implying that in every well there were always cells that did not commit to aggregation. Alternatively this lack of correlation is perhaps just indicative of natural variation in the exact numbers of cells plated into each well. In these experiments the presence of apyrase also induces the formation of clumps of cells. It was initially believed that these clumps were a result of premature expression of cell-cell adhesion molecules and this was measured using an already existing protocol. This protocol did not, however, detect any impact of addition of apyrase. The experiment did, however, seem able to detect changes in cell-cell adhesion as demonstrated by adding EDTA and observing the rapid fall in Aggregation Quotient. These results, however, do not necessarily disprove the idea that these clumps are caused by cell-cell cohesion, it only proves that clumps did not form under these experimental conditions. Despite the fact that the cells were kept in the same buffer as in the submerged culture experiments, the environment was still very different. In the adhesion experiment the cells were rotated constantly. This prevents development occurring because the mixing ensures an even distribution of camp and camp-degrading PDE and prevents the formation of a gradient. If the clumps indeed consisted of cells adhering to each other then one could still get the results seen in these experiments. This would mean that Clump formation is dependent upon the expression of the developmental gene pathways which are triggered only once the camp waves are detected. It could be definitively confirmed 56

57 whether this was the case with an investigation into the mrna expression of the cell-cell adhesion genes, such as cada, during development in control conditions and with apyrase. An alternate explanation for the presence of the clumps could be that, rather than being cells sticking together, they are cells that have chemotaxed close to each other, causing a high density of cells in one area. Interestingly this is in some ways similar to the phenotype predicted for situations in which there is an increased amount of secreted camp-degrading phosphodiesterase compared to the membrane bound form. In this scenario, the creation of too many camp waves cause constructive interference and as a result many competing aggregation centres form (Palsson, 2009). It is not being suggested that the same effect is occurring in these experiments. In the simulations there this caused a in the number of aggregates formed, which was not the case in these experiments, but something similar could be occurring with many cells coming together too quickly and uncoordinatedly and then having to re-sort themselves afterwards before they can form streams. It is unknown whether the cells treated with apyrase form clumps when they develop on agar, as the individual cells cannot be visualised in these experiments. Perhaps a more powerful microscope could resolve this problem. If indeed it turned out that clumps did not form on agar, even though cells which developed on agar experienced the delay, it might lend credence to the idea that clumps are a separate phenotype from the delay. 57

58 Chapter 4 Investigating Ecto-ATPases And Their Role In Regulating Development In Dictyostelium discoideum 4.1 Aim The luciferase assay had shown the presence of ATP in the extracellular medium during development and suggested that artificial breakdown of ATP and ADP by apyrase impacted the developmental processes. This implied the existence of a purinergic signalling system involved in the regulation of Dictyostelium discoideum development. So far this hypothetical system had only been investigated by observing the phenotypes when ATP/ADP signalling was prevented. In this chapter investigations were begun into ascertaining how this system is regulated under normal conditions. As of the time of writing there are no known ATP/ADP secretion or detection systems within Dictyostelium discoideum, but it is known to have methods of breaking down extracellular ATP to ADP. This is, therefore, the most promising place to start investigation of this hypothetical purinergic signalling system. In this chapter experiments were carried out to measure Dictyostelium discoideum s natural capability to degrade ATP and whether developmental phenotypes could be induced by interfering with this Introduction There is evidence for the existence of a Dictyostelium discoideum cell surface enzyme which normally keeps the extracellular ATP levels in a steady state (Parish and Weibel, 1980). Experiments carried out by Parish and Weibel show that it is sensitive to Suramin, a common ecto-atpase inhibitor which also has an inhibitory effect on P2 receptors (Chen et al., 1996), and to the local concentration of Mg 2+ ions. If a way of measuring this protein s activity could be devised, then it would be possible to deduce the concentrations of these factors that would affect the cells in the submerged culture conditions. This would allow more fine-tuned interference with the ATP signalling system. Initially a modification of a phosphate liberation assay used to detect ecto-atpases in Acanthamoeba (Sissons et al., 2004) was used to determine whether the cells have their own ecto-atpase activity. Some of the modifications made when the assay was used in other organisms (Feng et al., 2011) were also taken into account during experimental design. This experiment works by measuring the changes in phosphate concentration following the addition of exogenous ATP, using the assumption that the phosphate is produced from ATP breakdown. It is assumed that if a greater rate of Phosphate liberation occurs in the presence of cells than in buffer alone, it is indicative of ecto-atpases breaking down ATP quicker than the background rate. Despite the fact that Sissons, Feng and Ludlow all take such results as evidence of ATPase activity, this experiment cannot differentiate between ATP breakdown and other sources of phosphate release. If the assay gave results indicative of the activity of Dictyostelium discoideum ecto-atpases, then the next stage would be to try and identify their inhibitors as a way of characterising them. It was not immediately assumed that common ecto-atpase inhibitors would be effective, due to the fact that other elements of this hypothetical ATP signalling pathway, such as the P2X-like receptor, have already been shown to not be affected by common inhibitors (Ludlow et al., 2009). If ecto-atpases could be shown to be present, it could be inferred that their function is to break down extra-cellular ATP. The rate of this breakdown was earlier implied to be critical for ensuring the 58

59 correct timing of developmental events and the makeup of fruiting bodies. This could be tested by treating the cells with inhibitors of the ATPase activity and observing whether this caused an effect on development. As the addition of extra ATPases (apyrase) caused a delay in stream formation, it was anticipated that inhibition of ATPases would cause an effect as well. One could be tempted to assume that ATPase inhibitors would provoke the opposite effect to apyrase, but biological systems are highly complex and calibrated precisely and it was far more likely that any human interference of the ATP signalling system would result in negative effects on the cells. 4.3 Phosphate Liberation Phosphate Liberation Measuring Change In Absorbance In Control Conditions In these experiments wells of a 24 well plate were set up containing either 5X10 5, 1X10 6 or 0 cells in 1ml of Phosphate Liberation Buffer. These plates were left for 90 minutes to allow cells to adhere to the sides of the wells and condition the media with secreted compounds. Then 0.5mM ATP, dissolved in Phosphate Liberation Buffer, was added and the wells were gently shaken for the next 6 minutes while samples of the medium were taken. Absorbance of light with a wavelength of 600nm through the supernatant of these samples was measured using a spectrometer. A change in the absorbance of light is indicative of a change in phosphate concentration and this is often assumed to be a result of ATPase activity (Sissons et al., 2004). A calibration curve of phosphate concentration was not drawn for these experiments. This is because, while the results within a single experiment remained with a similar range, there was large variation in the readings gained between different experiments. The most likely explanation for this is that the buffer had to be freshly made up for each experiment. Slight variations in buffer composition, as well as similar slight variations of the concentration of ATP added to this buffer daily, gave very different readings, even though the pattern of data remained the same. This means that the data for these experiments could not presented as a change in phosphate ion concentration, but is instead given as the change in absorbance. This large variation in the numbers gained between experiments also made initial attempts at interpreting the data difficult. In order to minimise this variation between experiments, each data point is shown as the difference between the absorbance at that specific time and the absorbance of a sample taken from the same well at time zero. As this would render each data point at time zero as having a value of zero with no variation, the data points for this time point have not been shown in the following graphs and only the data obtained after three and six minutes is displayed. 59

60 Change in Absorbance at 600nm since time zero 0.12 * 0.1 * 0.08 * * 3 minutes 6 minutes Control 1X10 6 cells/ml 2X10 6 cells/ml Figure 26 There Is Significant Change In The Absorption Of light From Samples Taken From Wells Containing Cells After Addition Of Exogenous ATP Between Time Zero And Either Three Minutes Or Six Minutes. Data at three minutes is significantly different in each of the samples from wells containing cells from the control wells at the same time point (p<0.05). There is no statistical difference at six minutes between the wells containing cells and the control that do not, but the wells containing cells still have significantly more absorbance (p<0.05) than the same well at time zero. Data points taken at 3 and 6 minutes. The error bars represent standard error of the mean of each data point, N=7 A paired T-Test of the values obtained at time zero and three minutes after addition of the 0.5mM ATP (Figure 26) found significant difference (P<0.01) in the absorbance of 600nm light in the samples that came from wells containing cells. There was no significant difference found in the readings obtained between these two time points in the control wells. There was also a significant difference (P<0.05) in the absorbance of 600nm light in the samples that came from wells containing cells between time zero and six minutes. Again there was no significant difference between readings obtained between these time points in the control wells. There was also statistical difference in the change in absorbance at three minutes between the wells containing cells and the control (p<0.05 in both instances). Interestingly there was no significant difference between the absorbance in wells containing cells and the control at 6 minutes, and indeed by looking at the graph one can see at both cell densities the change in absorbance since time zero has fallen to be at a similar level to the control by this time. A change in absorbance is indicative of a change in phosphate ion concentration (Feng et al., 2011). The most likely source of phosphate ions in this situation would be from the breakdown of the 60

61 Change in Absorbance at 600nm since time zero exogenous ATP. The significant difference observed at three minutes between the change in absorbance in the wells containing cells and those that do not could therefore be indicative of ATPase activity. This fails to account for the fact that there is no significant difference observed at six minutes Phosphate Liberation Measuring Change Of Absorbance In The Presence of Suramin Unfortunately there are no known specific inhibitors which only impact ecto-atpases (Knowles, 2011). Suramin is known to inhibit both mammalian ecto-atpases (Chen et al., 1996) and Acanthamoeba ecto-atpases (Sissons et al., 2004), so it seemed a suitable first choice for attempting to ascertain a compound that would inhibit the Dictyostelium protein. The previous experiment was repeated, but this time each well also contained 250µM suramin. This concentration of suramin was chosen as it was known to be sufficient to inhibit the ecto-atpases of Acanthamoeba, the most similar organism to Dictyostelium upon which this experiment had already been carried out minutes 6 minutes Control 1X10 6 cells/ml 2X10 6 cells/ml -0.1 Figure 27 In The Presence Of Exogenous Suramin, Exogenous ATP Causes No Alteration In Absorbance At 600nm In The Six Minutes After Addition. Addition of ATP (0.5mM) in the presence of suramin (250µM) causes no significant difference in absorbance of light at a wavelength of 600nm. Data points taken at 3 and 6 minutes. The error bars represent standard error of the mean of each data point, N=6 In these experiments there was no significant difference found in the change of absorbance at zero minutes and either three or six minutes after addition of ATP in any of the wells (Figure 27). There was also no significant difference observed in the change in absorbance in samples from wells containing cells and control wells. It had previously been suggested that, in the absence of suramin, cells cause an increase in phosphate ion concentration three minutes after 0.5mM exogenous ATP is added. This experiment indicates that, whatever was responsible for the significant differences observed in Figure 26 is inhibited by the addition of 250µM suramin. 61

62 Change in Absorbance at 600nm since time zero It is unfortunate that, due to the labour intensive nature of these experiments, it was impossible to carry them out for longer than six minutes in order to ascertain whether one would ever observe a change in absorbance ATPase Activity Measuring Change Of Absorbance In The Absence Of Mg 2+ Ions Many ATPases are dependent on Mg 2+ ions for their operation and these are present in both the phosphate liberation buffer and the development buffer. As well as testing if inhibition of ecto- ATPases by removal of suramin elucidated the data it was also tested whether removing Mg 2+ ions from the buffer would have similar results. This time the 2mM MgCl was left out of the buffer preparation. It was hoped that the Mg 2+ ions would not prove necessary for the MLG-phosphate reaction and this did appear to be the case * * minutes 6 minutes Control 1X10 6 cells/ml 2X10 6 cells/ml Figure 28 Removal Of Mg 2+ Ions From The Reaction Buffer Decreases The Rate Of Ecto- ATPase Activity. Addition of ATP (0.5mM) with MgCl left out of the reaction buffer causes a significant difference in absorbance of light at a wavelength of 600nm between samples taken from the wells containing cells at time zero and the same wells at six minutes. No significant difference was observed in the absorbance of light between samples taken from wells at time zero and three minutes in any of the wells. Data points taken at 0,3 and 6 minutes. The error bars represent standard error of the mean of each data point, N=6 A paired T-Test found a statistically significant difference in the absorption of light at 600nm six minutes after the addition of ATP (Figure 28) in the cells containing wells. No such significant difference was found at three minutes, implying that the absence of Mg 2+ ions prevents this change in absorbance from happening as quickly as it did in the initial experiments from section As change in absorbance is linked to phosphate release and it has been hypothesised that this phosphate release is a result of ATPase activity, these results would imply that ATPase activity is partially, but not completely, inhibited in the absence of Mg 2+ ions. It is important to note, however, that at six minutes no significant difference was observed between the change in absorbance in samples taken from wells containing cells and the control wells. This seems unusual and could be explained by the very large error bars found in this experiment. This is in turn probably a result of the alterations made to the phosphate liberation buffer. More experiments 62

63 would need to be carried out before it could be determined if this partial inhibition exists or if the results seen here are an artefact of the experimental procedure. 4.4 The Effect Of Suramin On Dictyostelium Development In Submerged Culture Suramin was capable of inhibiting the increase in phosphate ion concentration which had been observed when 0.5mM ATP was added to Dictyostelium discoideum. One possible explanation suggested for this was that it was inhibiting ecto-atpase activity. It had already been implied in Chapter 3 that increasing the rate of extracellular ATP breakdown, through addition of apyrase, causes a delay in the initiation of streaming. It was, therefore, speculated that if suramin did inhibit Dictyostelium ecto-atpases then it would possibly also have an effect on the timing of stream formation as it would lead to a decrease in the rate of extracellular ATP breakdown. In order to investigate this, Submerged Culture experiments were set up as in Chapter 3, but instead of adding apyrase, 250µM suramin was added. 63

64 Figure 29 No Effect Of Suramin On Development. Development of AX4 cells at a density of 5x10 5 cells/ml in cultures treated with 250µM of suramin or an equivalent volume of water over a period of 24 hours. Cells start to form early streams at 5.5 hours and by 6.5 hours streaming is fully underway. By 24 hours aggregates have formed. Time is given in hours on the left of the image. Images are of the most developmentally advanced cells found within the well. 64

65 Aggregates formed The addition of 250µM suramin had no observable effect on the cells development (Figure 29). As in the apyrase experiments the amount of aggregates formed in each well containing suramin treated cells was expressed as a percentage of the number of aggregates formed in the paired control well. There was no difference in the number of aggregates formed by cells developing in the presence of suramin (Figure 30). In submerged culture no difference could be detected between control and suramin treated cells Control 250µM Suramin Figure 30 Treatment With Suramin Caused No Significant Difference In The Number Of Aggregates Formed By Dictyostelium. Number of aggregates present in wells containing control or suramin (250µM) treated Dictyostelium seeded at 5X10 5 cells/ml following completion of development. The error bar represents standard error of the mean of the data, N=3, 2 wells/experiment 65

66 4.5 The Effect Of Suramin On Dictyostelium Development On Agar The effect of 250µM suramin upon development on agar was also investigated. Even though no change in timing had been observed in submerged culture, the apyrase experiments had indicated that the ATP signalling system is involved in more than just changing the timing of development. Figure 31 Suramin Has No Effect On The Timing Of Dictyostelium Discoideum Development On Agar. Development of AX4 cells seeded at a density of 1x10 4 cells/cm2 on solid non nutrient agar treated with 250µM of suramin or an equivalent volume of molecular-grade water at 19 and 43 hours after addition to the plate. At 19 hours cells in all the wells had formed mounds. By 43 hours all the plates contained fruiting bodies. There was no difference observed in the timing of the developmental stages between either the control and suramin treated cells (Figure 31). Furthermore, there was no significant difference in the number of fruiting bodies present in the control wells and those containing cells treated with suramin (Figure 32). When the fruiting bodies were spread on bacterial lawns in order to ascertain the number of viable cells within them, however, it was found that there was a significant difference (p<0.05). The fruiting bodies formed from the suramin treated cells produced on average 180% as many colonies (Figure 33). 66

67 Colonies produced Fruting Bodies Formed Control 250µM Suramin Figure 32 - Suramin Has No Effect On The Number Of Fruiting Bodies Formed. Mean number of fruiting bodies formed by Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar with or without suramin (250µM). Each data point is expressed as a percentage of the number of fruiting bodies in the paired control well. The error bar represents standard error of the mean the data set, N=7, two wells/experiment. 250 * Control 250µM Suramin Figure 33 - Evidence For Increased Number Of Viable Cells In Fruiting Bodies Developed In The Presence Of Suramin. Number of colonies produced by 0.5ml of the mixture of a fruiting body head and 4ml of Klebsiella when grown upon an agar plates. Spores produced from Dictyostelium seeded at a density of 1x10 4 cells/cm 2 on solid non nutrient agar with or without suramin (250µM). Each data point is expressed as a percentage of the number of colonies produced by the relevant control. N=29 plates of control and 21 of suramin. The error bar represents standard error of the mean the data set. 67

68 4.6 - Conclusions This chapter presents evidence for the existence of endogenous ecto-atpases in Dictyostelium discoideum and attempts to ascertain what role they play in development. When 0.5mM ATP is added to cells the local phosphate ion concentration increases. It is believed that this is a result of cell surface ecto-atpases breaking down this ATP. This experiment cannot, however, rule out the possibility that cells are releasing extracellular phosphate ions in response to the ATP. It can also not be shown that ATP breakdown is carried out by ecto-atpases and not ATPases secreted by the cells into the medium. It is important to remember, however, that the presence of these ecto-atpases had already been shown by Parish and Weibel in 1980 and that in 2008 Ludlow used the same method used here to measure ecto-atpase activity without mention as to a possible other source of this phosphate. Furthermore, the fact that suramin, a common inhibitor of ecto-atpases, prevents this observed increase in phosphate concentration and alters the number of viable cells in fruiting bodies, which Chapter 3 suggests is influenced by ATP concentration, provides valuable indirect evidence that it is ecto-atpase activity that is being observed in these experiments. Six minutes after addition of the exogenous ATP there is no longer a significant difference between the absorbance of samples from the wells containing cells and those that do not. This would imply that the cells are absorbing the phosphate that is produced. Phosphate plays an important role in a variety of Dictyostelium cell functions (Zhang et al., 2005) and it seems feasible that the cells would not waste this source of it. It is unfortunate that the lack of a phosphate concentration calibration curve makes it difficult to determine whether it is indeed a significant amount of phosphate that is produced and taken up. Mammalian ecto-atpases require the presence of Mg 2+ ions to operate, with an optimum concentration between 1-5mM. In fact, removal of Mg 2+ ions by means of EGTA normally completely inhibits ecto-atpase activity (Kukulski et al., 2005). Figure 28 suggests that this might not be the case for the Dictyostelium ecto-atpases as, although phosphate ion concentration decreases in the buffer without Mg 2+ ions, it is not completely stopped. It is important to note, however, that while the buffer was made up without Mg 2+, no attempts were made to keep it out of the extracellular medium and EGTA was not added. It is already known that the other cation necessary for ecto-atpase activity in mammals, Ca 2+ (Kukulski et al., 2005), is released extracellular by Dictyostelium discoideum (Sivaramakrishnan and Fountain, 2012). It seems possible that Dictyostelium could also release Mg 2+ into the medium, the lower rate of activity being a result of the lower concentration caused by the cells being the only source of Mg 2+. Indeed, it is known that the cells release their Ca 2+ ions in response to ATP and it seems plausible that one of the reasons is to increase the activity of the ecto-atpases which break down this ATP. There are two families of cell surface proteins which break down ATP. Ecto-ATPases break down ATP in preference to ADP, whereas the closely related ecto-atp diphosphohydrolases break down both ATP and ADP equally (Zimmermann et al., 1998). The protein observed by Parish and Weibel in 1980 appeared to break down ATP over ADP implying that it was only an ecto-atpase. The phosphate liberation experiments above do not differentiate between the two and, in fact, the apyrase used in the development experiments was of the type that would break down both nucleotides equally. The phenomenon observed did, however, react to suramin and Mg + in the same way as the one observed by Parish and Weibel, again implying that the experiment is measuring the ecto-atpase s activity. 68

69 As stated earlier, there are no specific inhibitors of ecto-atpases (Knowles, 2011), so there is a worry that the results seen from the introduction of suramin and removal of Mg + represent responses of other cellular pathways rather than inhibition of the ecto-atpases. In both treatments, however, there was a lack of significant change in phosphate concentration during the first three minutes, which implies that the same pathways are being affected in both. This is likely to be the ecto-atpase as it is unlikely that both treatments would have similar unknown side effects to a hypothesised effect. One of the known side effects of suramin is that it can inhibit P2X receptors and perhaps one might have expected a similar reaction in the suramin treated cells as in those treated with apyrase. Apyrase is, after all, assumed to cause its effect by preventing cells from detecting ATP and ADP, removing it before it can activate receptors. This is not the case, however, and indeed the existing work on Dictyostelium discoideum s responses to nucleotides indicate that, although its method of ATP/ADP detection is similar to P2X in many ways, it is not suramin sensitive (Ludlow et al., 2008). Given that the presence of exogenous ATP was demonstrated in chapter 3, it would be reasonable to assume that inhibition of the cell s natural ecto-atpases by suramin would raise the extracellular ATP levels. The major obstacle preventing the experimental confirmation of this is that suramin inhibits luciferase (Tatur et al., 2007) and so the assay used throughout this project to measure the presence of ATP would be unsuited for such experiments. Without this one cannot conclusively state whether the ATP levels do indeed rise in the presence of suramin or whether suramin actually impacts ATP breakdown in development buffer. The fact that one observes an alteration in the number of viable cells present within a fruiting body indicates that adding suramin during development causes interference with a system which Chapter 3 had suggested is regulated by ATP. Chapter 3 suggests that removal of exogenous ATP by the ATPase apyrase causes a delay in stream formation, but these results imply that inhibiting the breakdown of this ATP has no effect on the timing. This indicates that there is not a direct correlation between rate of ATP concentration and the initiation of streaming. One possible explanation could be that there is an ATP threshold that must be met for normal streaming to occur. If the ATP concentration falls below this level (for example, due to addition of apyrase) then the timing of streaming is impaired. The cells are indifferent, however, to the concentration either above or below this level. This accords with established knowledge of Purinergic receptors in other organisms where it has been shown that the ATP concentration must be a certain level for cation influx to occur (Baricordi et al., 1996). While it is important to remember that the ATP response of Dictyostelium discoideum does not correspond to any currently known purinergic receptor, it seems logical that it would have many of the same features. An alternative explanation would be that, although the response to ATP is concentration dependent, under control conditions the amount of ATP released gives the maximal response and the cells are physically incapable of developing any faster. Below the concentration found in control conditions, however, the rate of stream formation may be proportional to ATP concentration. Experiments into adjusting the concentration of apyrase added to the cells and comparing the timing between these would be a useful way of determining which hypothesis is more likely. That fruiting bodies formed from the suramin treated cells were able to produce more colonies than the control ones, despite there being no difference in the number of fruiting bodies themselves, lends credence to one theory postulated in the previous chapter. Interfering with the ATP signalling system, rather than changing the number of cells present in the fruiting bodies, could instead alter the number of these cells that are viable. It may be that the cell number remains the same within these fruiting bodies, but the number detectable during the bacterial lawn experiments changes. This hypothesis has 69

70 the advantage of not raising the question of where the additional cells within the fruiting bodies treated from suramin came from, as they would also be present under control conditions. One should bear in mind that, unlike in the apyrase experiments, suramin did not appear to alter the number of unaggregated cells following development. This would imply that a similar percentage of the cells within each well have entered the fruiting bodies in those containing cells treated with suramin as those under control conditions. This hypothesis does, however, raise the question of what unviable cells would be doing within the fruiting bodies under control conditions. Unfortunately the fact that unviable cells are by their very nature unviable, makes it difficult to test for their presence short of cracking open the fruiting bodies and physically counting the cells. This is unfeasible. Some of the experiments in this project produced fruiting bodies containing as few as 100 viable cells: an experiment where 25 colonies formed from 0.5ml of KA/Spore mix had a mean of 200 cells in the 4ml of KA into which the two spores had been initially suspended. The haemocytometer used in this lab loses accuracy when counting fewer than 5x10 5 cells/ml and has to be loaded with 10µl for each reading. Even assuming that one did not make more volume than required, one would still need to use 50 fruiting bodies for each count, even without taking into account the need for repeats and observing the cells under the different treatments. 70

71 Chapter 5 - Bioinformatics 5.1 Aim This chapter outlines the bioinformatic techniques used in an attempt to discover the genes responsible for the ATPase activity observed in Dictyostelium discoideum. 5.2 Introduction The experiments carried out throughout this thesis imply that extracellular ATP is used by Dictyostelium discoideum cells to coordinate their development. Knowledge of the genes involved would allow the creation of mutants, which could be better used to explain how this ATP signalling system works. The initial discovery of Dictyostelium P2X receptors was made using bioinformatic techniques (Fountain et al., 2007). One would assume that if the ATP receptor which the cells were using in these experiments was easily detectable by bioinformatics assays then it would have been identified at this point. The cells, however, have been confirmed to possess ecto-atpases (Parish and Weibel, 1980), and it is suggested in Chapter 4 that inhibition of these interferes with this ATP signalling system, as does adding extra endogenous ATPases. Although detection of this ATPase would not directly reveal the receptor, it would allow further investigation into modulating the pathway responses. 71

72 5.3 Overview Of The Process Used Collect sequences of known ecto-atpases Identify sequences conserved between them Search against Dictyostelium genome using BLAST Find transmembrane topology of the resulting Dictyostelium proteins Remove genes that do not have the right transmembrane topology Identify domains of remaining proteins, where possible 72

73 5.4 Conserved Sequences The amino acid sequences of the known Ecto-ATPases found in various other species were gathered from the PubMed database. For convenience in cataloguing them, each gene was referred to by its Gene Index, or gi, number. Arabidopsis thaliana- gi nucleoside triphosphate diphosphohydrolase 3 gi nucleoside triphosphate diphosphohydrolase 4 gi nucleoside triphosphate diphosphohydrolase 5 gi nucleoside triphosphate diphosphohydrolase 6 gi nucleoside triphosphate diphosphohydrolase 7 Bos taurus - gi ectonucleoside triphosphate diphosphohydrolase 1 gi ectonucleoside triphosphate diphosphohydrolase 1 gi ectonucleoside triphosphate diphosphohydrolase 5 gi ectonucleoside triphosphate diphosphohydrolase 8 gi ectonucleoside triphosphate diphosphohydrolase 8 Danio rerio - gi ectonucleoside triphosphate diphosphohydrolase 1 Drosophila melanogaster - gi NTPDase 6 gi ectonucleoside triphosphate diphosphohydrolase 2 gi ectonucleoside triphosphate diphosphohydrolase 3 gi ectonucleoside triphosphate diphosphohydrolase 4 gi ectonucleoside triphosphate diphosphohydrolase 6 Gallus gallus - gi ecto-atp-diphosphohydrolase gi ectonucleoside triphosphate diphosphohydrolase 2 gi ectonucleoside triphosphate diphosphohydrolase 8 Homo sapiens - gi ectonucleoside triphosphate diphosphohydrolase 1 gi ectonucleoside triphosphate diphosphohydrolase 2 gi ectonucleoside triphosphate diphosphohydrolase 3 gi ectonucleoside triphosphate diphosphohydrolase 6 gi ectonucleoside triphosphate diphosphohydrolase 7 gi ectonucleoside triphosphate diphosphohydrolase 8 gi Ectonucleoside triphosphate diphosphohydrolase 8 gi small cell lung carcinoma ecto-atpase gi liver ecto-atp-diphosphohydrolase 73

74 gi Crystal Structure & Domain Rotation Ntpdase1 Cd39 Leishmania major - Lolium perenne - gi putative NTPDase gi Apyrase Mus musculus - gi ectonucleoside triphosphate diphosphohydrolase 1 gi ectonucleoside triphosphate diphosphohydrolase 2 gi ectonucleoside triphosphate diphosphohydrolase 4 gi ectonucleoside triphosphate diphosphohydrolase 5 gi ectonucleoside triphosphate diphosphohydrolase 7 gi ectonucleoside triphosphate diphosphohydrolase 8 Pongo abelii - gi ectonucleoside triphosphate diphosphohydrolase 7 Rattus norvegicus - gi ectonucleoside triphosphate diphosphohydrolase 1 gi ectonucleoside triphosphate diphosphohydrolase 3 gi ectonucleoside triphosphate diphosphohydrolase 6 gi ectonucleoside triphosphate diphosphohydrolase 8 Salmo salar - gi Ectonucleoside triphosphate diphosphohydrolase 2 gi Ectonucleoside triphosphate diphosphohydrolase 2 gi Ectonucleoside triphosphate diphosphohydrolase 6 Sus scrofa - gi ectonucleoside triphosphate diphosphohydrolase 1 Trifolium repens - gi Apyrase gi Apyrase Trypanosoma brucei - gi TPA_exp: c Xenopus laevis - gi ecto-nucleosidase triphosphate diphosphohydrolase 3 gi ectonucleoside triphosphate diphosphohydrolase 4 gi ectonucleoside triphosphate diphosphohydrolase 6 gi ecto-nucleosidase triphosphate diphosphohydrolase 6 gi ectonucleoside triphosphate diphosphohydrolase 7 Xenopus (Silurana) tropicalis - gi NTPDase5 gi NTPDase6 74

75 Having gathered the various sequences, the next step was to find regions that were highly conserved between them. The reasoning being that if the gene that fulfils a function has the same sequence of amino acids, even in species that are only very distantly related, then this region is likely to be highly important for the carrying out of this function (Alberts, 2008). One cannot, however, just line up two proteins next to each other in order to find these conserved sequences. The millennia over which the genes have evolutionarily diverged are more than enough time for multiple amino acid insertions, inversions and deletions to occur, making the distance between these conserved regions vary widely between organisms. In order to find these conserved regions, one must carry out a sequence alignment. A computer program is used to compare many sequences and line them up on the basis of the incidence of certain amino acid sequences at certain positions along the protein. One such program is ClustalW (Thompson et al., 1994). The amino acid sequences of the proteins, which had been identified by PubMed, were entered into ClustalW and the conserved regions identified. This output is displayed over the next two pages in Figure

76 76

77 Figure 34 - Sections Of The Clustalw Output (Larkin et al., 2007) With The Regions That Were Highly Conserved Marked And Numbered 77